University of Southern California Dissertations and Theses

Advancing Redwood City's bicycle infrastructure through a geodesign workflow

(USC Thesis Other)

Advancing Redwood City's bicycle infrastructure through a geodesign workflow

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Advancing Redwood City’s Bicycle Infrastructure Through a Geodesign Workflow

by

Erik Siebren Huisman

A Thesis Presented to the
FACULTY OF THE USC DORNSIFE COLLEGE OF LETTERS, ARTS AND SCIENCES
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY)

August 2023

Copyright 2023 Erik Siebren Huisman
ii

To my family, my friends, and my hometown
iii
Acknowledgements

I am grateful to my thesis advisor, Elisabeth Sedano, for providing me with the guidance and
insight to complete this project. I am also grateful for the help provided to me by Leilei Duan,
Guoping Huang, Darren Ruddell, and the rest of the USC Spatial Sciences Institute. I also want
to thank Jackie Campos and Rafael Avendaño of Redwood City Together for their help with
coordinating the workshop as well as Malahat Owrang from the City of Redwood City for her
help with coordinating the workshop and providing me with assistance regarding the acquisition
of spatial data. A big thank you also goes out to my friends and family for proofreading my work
and providing me with valuable feedback and unwavering support. Lastly, I am appreciative for
the support and confidence in me shown by the Redwood City Safe Routes to School task force.
iv
Table of Contents
Dedication ...................................................................................................................................... iii
Acknowledgements ....................................................................................................................... ivi
List of Tables ................................................................................................................................. vi
List of Figures ............................................................................................................................... vii
Abbreviations .................................................................................................................................. x
Abstract .......................................................................................................................................... xi
Chapter 1 Introduction .................................................................................................................... 1
1.1 Research Goals.................................................................................................................... 2
1.2 Study Area .......................................................................................................................... 3
1.3 Cycling in Redwood City.................................................................................................... 7
1.3.1 Existing Cycling Conditions ...................................................................................... 7
1.3.2 Walk Bike Thrive Initiative ..................................................................................... 10
1.4 Motivation ......................................................................................................................... 13
1.5 Thesis Overview ............................................................................................................... 15
Chapter 2 Related Literature ......................................................................................................... 16
2.1 Benefits of Biking ............................................................................................................. 16
2.2 Social Factors and Bikeability .......................................................................................... 19
2.3 GIS Methodologies and Bikeability.................................................................................. 22
2.3.1 Variables Used to Quantify Bikeability ................................................................... 22
2.3.2 GIS Methodologies to Quantify Bikeability ............................................................ 27
2.3.3 Evaluating the Equity Impact of Cycling Infrastructure .......................................... 31
2.4 Urban Design Techniques for Improved Bikeability ........................................................ 32
2.4.1 Public Engagement .................................................................................................. 32
2.4.2 Policy Tools ............................................................................................................. 33
2.4.3 Physical Infrastructure ............................................................................................. 34
2.4.4 Geodesign as a Tool to Improve the Built Environment ......................................... 35
Chapter 3 Methods ........................................................................................................................ 36
3.1 Methods Overview ............................................................................................................ 36
3.2 Bikeability Quantification ................................................................................................. 37
3.2.1 Variable Choices and Data Preparation ................................................................... 38
3.2.2 Weighted Sum to Quantify Bikeability.................................................................... 58
3.2.3 Reclassifying Bikeability for Site Selection ............................................................ 59
3.2.4 Demographic Comparison ....................................................................................... 61
3.3 Workshop .......................................................................................................................... 62
3.3.1 Workshop Planning .................................................................................................. 62
3.3.2 Conducting the Workshop ....................................................................................... 70
v
3.3.3 Assessment of Workshop Results ............................................................................ 71
3.4 Equity Analysis ................................................................................................................. 76
3.5 Site Selection .................................................................................................................... 83
3.6 Proposed Project Design Concepts ................................................................................... 84
3.6.1 Site Investigation ..................................................................................................... 84
3.6.1.1 Current status assessment ............................................................................... 84
3.6.1.2 Demographics of the surrounding neighborhood ........................................... 85
3.6.1.3 Zoning designations of the surrounding neighborhood .................................. 86
3.6.2 Modeling .................................................................................................................. 86
Chapter 4 Results .......................................................................................................................... 93
4.1 Bikeability Quantification ................................................................................................. 93
4.1.1 Redwood City Bikeability ....................................................................................... 94
4.1.2 Redwood Shores Bikeability.................................................................................... 96
4.1.3 Proposed Projects’ Bikeability ................................................................................. 97
4.2 Workshop Findings ........................................................................................................... 98
4.3 Equity Impact Quantification .......................................................................................... 103
4.4 Site Selection .................................................................................................................. 105
4.5 Proposed Project Design Proposals ................................................................................ 107
4.5.1 Redwood Avenue ................................................................................................... 108
4.5.2 Vera Avenue .......................................................................................................... 114
4.5.3 Hudson Street ......................................................................................................... 121
Chapter 5 Conclusions ................................................................................................................ 128
5.1 Bikeability Quantification ............................................................................................... 128
5.2 Workshop ........................................................................................................................ 129
5.3 Equity Analysis ............................................................................................................... 130
5.3.1 Bikeability and Demographics in Redwood City .................................................. 131
5.3.2 Bikeability and Demographics in Redwood Shores .............................................. 134
5.4 Site Selection .................................................................................................................. 137
5.5 Urban Design Modeling .................................................................................................. 139
5.6 Overall Utility of Methods .............................................................................................. 141
References ................................................................................................................................... 143

vi
List of Tables
Table 1. Variables used in previous studies .................................................................................. 23
Table 2. Required data .................................................................................................................. 39
Table 3. Speed limit reclassification ............................................................................................. 41
Table 4. Zoning designation reclassification ................................................................................ 44
Table 5. Slope reclassification ...................................................................................................... 46
Table 6. Tree canopy reclassification ........................................................................................... 47
Table 7. Bike lane access reclassification ..................................................................................... 55
Table 8. Crashes reclassification ................................................................................................... 57
Table 9. Bikeability weights ......................................................................................................... 58
Table 10. Workshop survey questions .......................................................................................... 67
Table 11. Proposed projects mentioned during the workshop ...................................................... 73
Table 12. Public feedback scores for proposed projects ............................................................. 102
Table 13. Fifteen highest-scoring proposed projects in terms of prioritization .......................... 106

vii
List of Figures

Figure 1. Redwood City, California ................................................................................................ 4
Figure 2. Redwood City neighborhoods ......................................................................................... 5
Figure 3. Redwood City arterial roads ............................................................................................ 7
Figure 4. Existing bike lane network .............................................................................................. 8
Figure 5. Redwood City bike infrastructure classes ....................................................................... 9
Figure 6. Proposed bike lane network ........................................................................................... 11
Figure 7. Existing and proposed bike lane network ...................................................................... 13
Figure 8. Emissions of 13 modes of transportation in kg GGE/PMT (Dave 2010)...................... 19
Figure 9. Portland service area breaks .......................................................................................... 28
Figure 10. Portland bikeability scores .......................................................................................... 29
Figure 11. Input factors and bikeability for the Vancouver Metropolitan Area ........................... 30
Figure 12. Methodology overview ................................................................................................ 37
Figure 13. Bike lane access ModelBuilder layout ........................................................................ 52
Figure 14. Bike lane access merge ModelBuilder layout ............................................................. 54
Figure 15. Proposed project bikeability ModelBuilder layout ...................................................... 60
Figure 16. Workshop outreach flier. ............................................................................................. 65
Figure 17. Activity 1 Miro board .................................................................................................. 68
Figure 18. Activity 2 Miro board .................................................................................................. 69
Figure 19. Activity 3 Miro board .................................................................................................. 70
Figure 20. C/CAG Equity Focus Areas ........................................................................................ 77
Figure 21. SamTrans Equity Planning Areas ................................................................................ 78
viii
Figure 22. MTC EPCs................................................................................................................... 79
Figure 23. CAHPI polygons ......................................................................................................... 81
Figure 24. Equity score ModelBuilder layout ............................................................................... 82
Figure 25. OSM data download steps ........................................................................................... 88
Figure 26. Netherlands bike lane driveway infrastructure ............................................................ 91
Figure 27. Redwood City bikeability ............................................................................................ 94
Figure 28. Redwood City bikeability (excluding Redwood Shores) ............................................ 95
Figure 29. Redwood City bikeability and neighborhoods ............................................................ 96
Figure 30. Redwood Shores bikeability ........................................................................................ 97
Figure 31. Proposed project bikeability scores ............................................................................. 98
Figure 32. Proposed projects mentioned in the workshop ............................................................ 99
Figure 33. 500ft buffer around parks and schools ...................................................................... 100
Figure 34. Proposed project community feedback scores .......................................................... 103
Figure 35. Proposed project equity scores .................................................................................. 104
Figure 36. Proposed project prioritization scores ....................................................................... 105
Figure 37. The three proposed projects selected to receive design proposals ............................ 107
Figure 38. Redwood Avenue proposed project site .................................................................... 108
Figure 39. Amenities near the Redwood Avenue proposed project ........................................... 109
Figure 40. Zoning designations near the Redwood Avenue proposed project ........................... 110
Figure 41. Redwood Ave. illustration of 2023 (top) and as proposed (bottom) ......................... 111
Figure 42. Redwood Avenue redesign aerial view ..................................................................... 112
Figure 43. Redwood Avenue redesign perspective view ............................................................ 113
Figure 44. Vera Avenue proposed project site ............................................................................ 115
ix
Figure 45. Amenities near the Vera Avenue proposed project ................................................... 116
Figure 46. Zoning designations near the Vera Avenue proposed project ................................... 117
Figure 47. Vera Avenue redesign in StreetMix .......................................................................... 118
Figure 48. Vera Avenue redesign aerial view ............................................................................. 119
Figure 49. Vera Avenue redesign perspective view ................................................................... 120
Figure 50. Hudson Street proposed project site .......................................................................... 122
Figure 51. Amenities near the Hudson Street proposed project ................................................. 123
Figure 52. Zoning designations near the Hudson Street proposed project ................................. 124
Figure 53. Hudson Street redesign in StreetMix ......................................................................... 125
Figure 54. Hudson Street redesign aerial view ........................................................................... 126
Figure 55. Hudson Street redesign perspective view .................................................................. 126
Figure 56. Bikeability and zero car household percentage in Redwood City ............................. 132
Figure 57. Bikeability and median household income in Redwood City ................................... 133
Figure 58. Bikeability and non-white percentage in Redwood City ........................................... 134
Figure 59. Bikeability and zero car household percentage in Redwood Shores ......................... 135
Figure 60. Bikeability and median household income in Redwood Shores ............................... 136
Figure 61. Bikeability and non-white percentage in Redwood Shores ....................................... 137
Figure 62. Relationship between the designed projects and equity polygons ............................ 139

x
Abbreviations

CAD Computer-aided design
CAHPI California Healthy Places Index
C/CAG City/County Association of Governments
DEM Digital Elevation Model
EPCs Equity Priority Communities
GIS Geographic information system
IRB Institutional Review Board
MPH Miles Per Hour
MTC Metropolitan Transportation Commission
OSM OpenStreetMap
SKP2OSM SketchUp to OpenStreetMap Plug-In
US United States
USC University of Southern California
WBTI Walk Bike Thrive Initiative

xi
Abstract
In the United States, emission-releasing cars reign as the leading form of transportation among
citizens. Given the increasing effects of global climate change, it is critical that society finds
alternative solutions to travel. The use of geodesign, combining data-driven spatial analysis with
thoughtful design and community input, shows promise as an approach to better design
transportation infrastructure in the US. This thesis applies a geodesign methodology to propose
biking infrastructure improvements in Redwood City, CA. It first assesses existing bikeability as
of 2022 using a spatial analysis in a GIS. It finds that Redwood City has moderate bikeability
with potential for improvements that if implemented, will simultaneously help solve issues
related to economic, social, and transport inequity. The project next selects three specific street
segments in the city that could most benefit from improved biking infrastructure. It makes the
selection through a combined analysis of these bikeability results, assessments by local
stakeholder organizations of underserved areas, and community feedback gathered at a public
workshop organized by the author that focused on biking in the city. These three selected street
segments underwent a design process, resulting in models and renderings of what improved
cycling infrastructure could look like in Redwood City. This thesis ultimately serves as an
exemplar methodology that can be applied to other cities in the US to increase local bikeability
and improve long-term sustainability in terms of social equity and the environment.
1
Chapter 1 Introduction
Cycling is an efficient and sustainable method for humans to travel within a community, city, or
region. However, American city planners and municipalities tend to focus less on implementing
bicycle infrastructure and instead prioritize high-speed freeways, wide parkways, and sprawling
parking lots. American car-dependency poses multiple challenges for residents, such as the high
cost of car-ownership which burdens already-disadvantaged communities. Heavy reliance on
cars also poses challenges for the environment through the release of large amounts of carbon
dioxide and other greenhouse gases. In contrast, bike use is relatively inexpensive, correlates
with improved physical and mental health, and releases no greenhouse gasses. Despite the
obvious benefits of bike use, it is difficult to safely and consistently ride a bike to perform daily
tasks in many cities across the United States (US). Biking to work or the grocery store is
challenging because bicycle infrastructure is usually nonexistent, and what infrastructure does
exist is often nothing more than a simple line of paint on the road which offers cyclists zero
physical protection from passing cars. Also, many American communities lack the connectivity
to form efficient cycling networks. Cycling as a form of transportation in the United States (US)
can be improved by increasing bikeability, which is defined as “the extent to which an
environment is convenient and safe for cycling” (Reggiani et al. 2022).
This thesis focuses on bikeability in Redwood City, CA. Redwood City is a small city in
California’s Silicon Valley with a government that is interested in improving biking
infrastructure. As part of the city’s Walk Bike Thrive Initiative (WBTI), a long-term plan for
municipal urban improvements, the city has already proposed 132 cycling infrastructure
improvement projects (referred to as “proposed projects” throughout this thesis). This thesis
creates design concepts for three proposed projects. It selects these three by evaluating current
2
bikeability across the city, considering the relationship between the proposed projects and social
equity, and listening to community priorities in an author-led community workshop. The
remainder of this chapter describes the research goals, the study area, a discussion of present-day
cycling in Redwood City, and a summary of the project’s motivation.
1.1 Research Goals
This thesis project leveraged a geodesign methodology, combining spatial analysis,
stakeholder engagement, and urban design modeling to provide a roadmap for increasing
Redwood City’s bikeability.
The first goal of the project was to analyze and quantify bikeability in Redwood City.
This was performed using a weighted sum calculation in a geographic information system (GIS)
with six variables conducive to bikeability.
The second goal was to assess the city’s WBTI proposed projects in terms of social
equity. Each proposed project was evaluated in a GIS to calculate the extent to which it would
benefit communities that have historically been underserved in terms of social and economic
resources. This assessment leveraged the knowledge of four municipal stakeholder organizations
that have previously identified underserved communities using a variety of demographic metrics,
such as median household income and ethnicity.
The third goal was to involve Redwood City residents in the methodology and
advancement of cycling in Redwood City. To do this, a workshop was held to gain insight from
Redwood City residents and allow them to voice their concerns about the state of local
bikeability in 2023 as well as their wishes for the future of local bikeability.
3
The fourth goal was to select three of the city’s WBTI proposed projects to undergo a
design process. The selections were made by combining the results of the first three components
in a weighted sum calculation.
The fifth research goal was to create 3D designs of what improved infrastructure could
look like for the three selected proposed projects. This was done using SketchUp, a 3D modeling
software.
Overall, this project used a mixed-methods, interdisciplinary, and nonlinear approach to
address Redwood City bikeability and provide valuable insight to city decision makers. The
geodesign methodology used in this project can be applied to other cities where relevant and
thorough spatial data can be acquired to support efforts to improve bicycle infrastructure.
Increased bikeability across American cities will be a positive change for the environment and
public health and will also reduce equity gaps.
1.2 Study Area
The study area for this thesis is Redwood City, CA (Figure 1. Redwood City, Californi).
Redwood City is located in Silicon Valley on the San Francisco Bay Area Peninsula and is
equidistant between San Francisco and San Jose. Redwood City is bordered by Atherton to the
southeast, Woodside to the south, San Carlos to the northwest, and the unincorporated
communities of North Fair Oaks to the east and Emerald Hills to the west. The official city
boundary encompasses roughly 35 square miles, of which approximately half is marshland or
San Francisco Bay water. The legal limits of Redwood City include the community of Redwood
Shores, which is located about three miles northwest of downtown Redwood City. Redwood City
is a suitable study area for a geodesign project focused on advancing bikeability because of its
4
existing bike infrastructure, generally flat terrain, and government interest in improving the bike
network.

Figure 1. Redwood City, California

Following the tech boom of the late 1990s and the influx of businesses and residents to
Silicon Valley, Redwood City grew into a denser and more diverse city. According to the US
Census Bureau, the population of Redwood City was 84,518 in 2020. The residents live in a mix
of single and multi-family homes, with a majority of dense housing being located in the
following neighborhoods: Centennial, Downtown, Stambaugh-Heller, and Redwood Village
(neighborhoods 11, 12, 14, and 15 in Figure 2. Redwood City neighborhoods). Much of the
central, southern, and western areas of Redwood City feature suburban-style single family
5
homes. It is worth noting that Redwood City’s streets for the most part follow a grid-like
structure. Cul-de-sacs, which are detrimental to bikeability, are rare. According to US Census
data, approximately 41.1% of residents identify as White-non-Hispanic, 35.3% as Hispanic or
Latino, and 16.3% as Asian (United States Census Bureau 2022). In 2022, the median household
income was $134,287, and the poverty rate was 7.1% (United States Census Bureau 2022).
Redwood City is undergoing rapid change and growth and in April of 2023, was
designated as a Prohousing Community by California Governor Gavin Newsom. California cities
can apply to receive the Prohousing designation which if given, allows them to be eligible for
funding and incentives conditioned on the construction of affordable and sustainable housing as
well as updated infrastructure (California Department of Housing and Community Development
n.d.). Redwood City is one of 22 Prohousing communities in California (Iracheta 2023).
6

Figure 2. Redwood City neighborhoods

The terrain and urban layout of Redwood City make it naturally conducive to good
bikeability. Much of the city is relatively flat, however as one travels further west, the terrain
becomes hillier. Alameda de las Pulgas, a main arterial road that runs from the northwest to the
southeast (see Figure 3. Redwood City arterial roads), generally acts as a border between flat
land to the east, where a majority of residents live, and sloped land to the west. Downtown
Redwood City, boxed in by El Camino Real, Chestnut Street, Veterans Boulevard, and Brewster
Avenue (see Figure 3), is home to an abundance of dining, retail, and nightlife options and also
serves as a hub for business headquarters, many of which are involved in the technology sector.
The downtown area also features a CalTrain station, which offers rail connectivity between San
7
Francisco and San Jose, and a SamTrans bus depot, which offers bus connectivity within San
Mateo County.

Figure 3. Redwood City arterial roads

1.3 Cycling in Redwood City
While Redwood City does maintain a significant cycling network, the current
infrastructure is lackluster and does not provide safety or connectivity to cyclists. However, the
city has expressed a desire to improve cycling and has outlined these goals in the WBTI.
8
1.3.1 Existing Cycling Conditions
As of 2023, Redwood City does have a cycling network (Figure 4). The existing bike
lanes do not offer a high amount of connectivity and most of the infrastructure is made up of
Class II and Class III bike lanes which offer no physical protection to the rider and are limited to
painted markings on the street that indicate where cyclists should ride. Class I and Class IV bike
lanes provide physical protection for cyclists, but they are not common in Redwood City and are
not located in efficient places; they are generally located near the marshes and San Francisco
Bay where they provide good bike infrastructure for leisure riding, but not for commuting.

Figure 4. Existing bike lane network

9
Each instance of cycling infrastructure in Redwood City falls into one of four classes.
Class I bike infrastructure refers to bike trails that are fully separated from traffic and in most
cases, do not run parallel to roads (Figure 5, top left). The existing Class I lanes are concentrated
along Redwood City’s coast and support cycling as a form of recreation rather than
transportation. Class II bike infrastructure refers to bike lanes that share roadways with cars but
have their own space as indicated by painted white lanes and green lanes (Figure 5, top right).
Class II bike lanes offer no physical protection for cyclists. Class III bike infrastructure are roads
that have been designated as bike routes, but cyclists are expected to share the road with cars and
do not have their own space (Figure 5, bottom left). Class III bike infrastructure offers no
physical protection and the only indication that these roads are intended for bike use are the
stenciled bike sharrows painted on the street. Class IV bike infrastructure refers to bike lanes that
are fully separated from the street but run parallel to existing roads (Figure 5, bottom right). They
offer ample physical protection for cyclists. Figure 5 shows an example of each of the four bike
infrastructure classes in Redwood City.
10

Figure 5. Redwood City bike infrastructure classes

1.3.2 Walk Bike Thrive Initiative
In 2022, the city published the comprehensive WBTI, which aims to improve various
physical aspects of the city and its infrastructure by developing policy to create a safe, walkable,
and bikeable urban environment. The WBTI is a 263-page document that includes written plans,
maps, and statistics. One of the primary targets of the initiative is to improve bikeability, which
includes adding new bike lanes (Figure 6), creating bike boulevards which are physically
separated from streets, adding additional bike parking near desirable destinations, and improving
street connectivity. While this plan demonstrates the elected city officials’ awareness of the
utility of increased cycling in terms of public health and environmental sustainability, the
11
physical infrastructure and policy must be crafted efficiently and effectively to prove meaningful
and ultimately receive implementation.

Figure 6. Proposed bike lane network

The city has made clear that the full implementation of the proposal is not guaranteed,
and it is therefore critical that the most valuable proposed additions to the bike lane network are
identified and prioritized. To identify these proposed projects, GIS has been used to assess the
strength of bicycle infrastructure as it existed in 2022 and evaluate how future bike infrastructure
may resolve issues related to the equal allocation of transportation resources to historically
disadvantaged communities. Projects completed in communities that have been identified by
12
stakeholders as being historically underserved and faced with equity issues generally have a
higher likelihood of receiving regional and state funding. From a financial perspective, it is
imperative that those projects are prioritized to maximize the number of infrastructure
improvements made in the city, given the limited budget. The overall goal of this thesis project
was to design and execute a methodology that could assist city staff and council members in their
decision-making process for proposed projects listed in the WBTI, which would contribute to an
advanced bike lane network. If the proposed projects are fully implemented, cycling connectivity
would be high across the city and gaps in the current network would be filled, especially within
the eastern half of the city. In addition, fully protected routes would protrude from downtown
towards the southwest into the residential areas, making cycling a more attractive option for the
city’s residents (Figure 7). This thesis also directly supports one of the six main goals set by the
WBTI: “Invest in projects that support a resilient, equitable, and sustainable transportation
system” (City of Redwood City 2022).
13

Figure 7. Existing and proposed bike lane network
1.4 Motivation
As a primary mode of urban transportation, cycling shows immense promise as it is faster
than walking and better for the environment than driving. A cyclist is also 10 times less likely to
seriously injure or kill pedestrians when compared to an automobile driver (Wardlaw 2000).
Both of these facts make cycling a good option for local transit, especially for travel distances
under five miles (Qin et al. 2018). The US lags behind many countries in prioritizing safe
cycling; with its significant automotive culture, it has historically been rare for policymakers to
propose investing in cycling infrastructure instead of car-centric infrastructure. By ignoring
cycling, municipalities are depriving their citizens of the benefits of biking as a form of
14
transportation, such as lower rates of obesity and heart disease (de Hartog et al. 2010), fewer
greenhouse gas emissions (Dave 2010), less noise pollution (Pucher and Buehler 2008), as well
as the opportunity for improved mental health that biking provides (Olsson et al. 2013). Instead,
Americans are left to be content with driving most places and sitting in traffic for hours, just to
get to work.
Spatial science and GIS, especially in the context of geodesign, are becoming
increasingly recognized as powerful tools for urban planning. GIS has tremendous value when
making urban planning decisions and can effectively guide city layout and design practices
(Weimin and Milburn 2016). This project developed a methodology in which GIS plays a heavy
role. GIS was first used to quantify the bikeability of the study area. Additional GIS techniques
were then deployed to determine the impact of proposed bike lanes on issues related to equity in
terms of local access to efficient and safe transportation options. Similar studies have been
conducted in the past, albeit for different study areas, and their methodologies serve as
inspiration for the design of this thesis’ bikeability evaluation overlay. These studies have taken
place around the world such as in Nigde, Turkey (Olgun 2020), Barranquilla, Colombia
(Arellana et al. 2020), Vancouver, Canada (Winters et al. 2013), and Portland, United States
(McNeil 2011). This project was conducted on the Bay Area Peninsula, which resulted in a
unique output given the study area’s topography, demographics, and progressive ideology which
includes a strong promotion of bicycle transit and transit-oriented development. The output
metrics were then made accessible to policymakers who can use the information to guide
decisions regarding where to improve bicycle infrastructure, and how to best do it.
The author also has personal interest in this topic. As a Redwood City native, the author
has witnessed firsthand the lack of bikeable infrastructure in the city and has significant interest
15
in helping to improve it by providing policymakers with evidence to make their decisions. In
addition, as a dual-citizen of the United States and the Netherlands, the author has witnessed
firsthand how effective bike infrastructure in the Netherlands contributes to biking being a
feasible method of transportation for all. As a result, the author has a desire to draw upon some
of the urban design techniques found in the Netherlands as inspiration for cycling infrastructure
in Redwood City, doing so through the application of a geodesign methodology.
The ultimate goal of this work was to produce deliverables that both underline the
importance of bikeability in the United States and exemplify a methodology for evaluating and
designing bike infrastructure. The contents of this thesis will hopefully lead to more cities
adopting a pro-bicycle stance in terms of local transportation. It also highlights and encourages
the application of the field of geodesign, a discipline that shows immense value and promise as
an urban design tool.
1.5 Thesis Overview
The remainder of this thesis includes four chapters. The related literature chapter covers
topics including the health and environmental benefits of cycling, the relationship between social
factors and demographics and cycling, the utility of GIS in evaluating bikeability, and urban
design techniques that can be applied to improve bikeability. The methods chapter outlines the
techniques that were used to conduct the thesis and provides justifications for certain decisions
made during the process. The results chapter describes the outputs of the methodology while the
conclusions chapter provides commentary on the results and future implications of the research.

16
Chapter 2 Related Literature
This chapter reviews literature on urban cycling and bikeability. To form a basis on which the
goal of advancing cycling infrastructure is justified, resources that describe the benefits of biking
and literature that discusses the social and physical factors related to cycling, are reviewed. To
aid in the formulation of a methodology to quantify the bikeability of each street segment in
Redwood City, literature that employs geospatial methodologies to evaluate bikeability are
reviewed. Finally, studies on the relationship between urban design and bikeability are reviewed.
These reviewed literatures formed the groundwork for the methodology devised and deliverables
created for this project.
2.1 Benefits of Biking
Unlike alternative methods of transportation such as cars, motorcycles, and trains, biking
does not require an external fuel source; rather the human body provides the kinetic energy
required for movement. Biking is faster and more efficient than other human-powered forms of
transportation such as walking and running, as it allows humans to travel further distances in
shorter amounts of time.
Because biking requires active movement by the rider, it is inherently beneficial to one's
physical health. Cycling is a form of cardiovascular, aerobic exercise and can help riders
maintain a healthy weight as well as lower their risk of heart disease (de Hartog et al. 2010).
Regular cycling has also been associated with a lower cancer mortality and morbidity rate (Oja et
al., 2011). Furthermore, the health benefits of biking far outweigh the risks posed by the dangers
of engaging in cycling, such as the risk of accidents and exhaust inhalation (Götschi, Garrard,
17
and Giles-Corti 2014). The risks of cycling are even lower if fewer cars are on the roads and
cyclists are allotted bike pathways separate from the street (Monsere et al. 2014).
Regular cycling can improve mental health and boost one’s mood. Olsson et al. (2013)
surveyed 713 Swedish citizens on work commute, happiness, and demographics. The results of
the survey suggest the level of satisfaction with one’s work commute has a significant impact on
that individual’s well-being and that walking or biking are much more satisfying than driving or
taking public transportation. However, in the US the disparity between satisfaction from driving
versus biking to work is not as large as in Sweden (Olsson et al. 2013). This may be due to
automotive commutes being more common in the US. Still, this study overall suggests that
commuting via bike improves one’s mood more than driving.
Ma, Ye, and Wang (2020) conducted a broad literature review and determined that
additional studies have corroborated the conclusion that cycling, especially when used as a form
of transportation rather than for leisure, can improve mental health by reducing feelings of
anxiety and depression (Ma, Ye, and Wang 2021). In another study, Dill and Rose (2021),
interviewed 28 individuals in Portland, Oregon who had recently switched from driving to using
an electric bike to get to work. One individual claimed that by cycling, he was “able to turn the
worst part of the day, which is getting in the car and driving to work into the best part of the
day,” implying that he finds his bicycle commute to be enjoyable (Dill and Rose 2021). Bicycles
also produce practically zero noise pollution, making them a peaceful transit option (Pucher and
Buehler 2008). Less exposure to traffic noise may provide other health benefits, as a 2018 study
found that exposing mice to traffic noise increased stress and anxiety and led to quicker brain
impairment and cognitive decline in the rodents (Jafari, Kolb, and Mohajerani 2018).
18
In addition to providing physical and mental health benefits, cycling is a form of
environmentally conscious transportation that emits significantly less greenhouse gasses and
noise pollution than alternative forms of travel. The primary environmental benefit of cycling is
that bicycles do not require the use of environmentally detrimental fossil fuels, and instead rely
on renewable resources taken in by the rider (Pucher and Buehler 2008). A study in Shanghai
investigated the environmental impacts of a bike-sharing system in which users rent bikes, likely
reducing automotive reliance. The study determined that in 2016, the program saved 8,358 tons
of gasoline. It also determined that carbon dioxide emissions had been reduced by 25,240 tons
and nitrogen dioxide by 64 tons (Zhang and Mi 2018). In the United States, a predictive study
performed in Portland, Oregon estimated that if 15% of passenger miles traveled were completed
on e-bike, carbon dioxide emissions could decrease by 12% (McQueen, MacArthur, and Cherry
2020). This provides evidence that an increase in cycling can quell climate change.
The bicycle manufacturing process is still inherently bad for the environment, given the
need for extracted mineral resources and the emissions released by factories and during the
transportation of parts (Chan, Schau, and Finkbeiner 2019). Regardless, the amount of
greenhouse gasses emitted over a bicycles’ lifespan is significantly lower than that of cars. This
is largely due to the disparity in associated emissions when being used as a means of transport. A
2010 study was undertaken to compare the difference in greenhouse gas emissions between
different modes of transportation. The study investigated the emissions related to the fuel
production, infrastructure, maintenance, manufacturing, and operation of 13 different
transportation options. The amounts of emissions per each form of transportation can be seen in
Figure 8, which comes directly from the study (Dave, 2010). It was determined that the average
SUV emits 446 kilograms of greenhouse gasses per passenger-mile-traveled while bikes emit 33
19
kilograms of greenhouse gasses per passenger-mile-traveled (Dave 2010). Lastly, bikes require
much less space than cars, suggesting that an increase in bike ridership could lead to a decrease
in construction of large roads and parking garages which would eliminate the large amounts of
embodied carbon associated with those types of infrastructure (Pucher and Buehler 2008).

Figure 8. Emissions of 13 modes of transportation in kg GGE/PMT (Dave 2010)

2.2 Social Factors and Bikeability
Social factors influence the convenience and safety of biking and a number of studies
have been conducted to identify specific relationships. Lindsey performed spatial analysis to
estimate the likelihood of a crash at all intersections and mid-blocks in Minneapolis based on
historic crash data (Lindsey et al. 2019). The output was then compared with demographic data
20
and revealed that a higher crash risk was related to areas with a lower average income, a
majority-minority population, and proximity to primary arterial roads.
These findings have been observed in other American cities, such as Los Angeles where
the risk of a crash involving a pedestrian was higher in areas with high Hispanic populations
(Loukaitou-Sideris 2007). The study first identified a correlation between high poverty rates and
high Hispanic populations, aggregated by census tract. It then concluded that because of lower
economic status, a greater percentage of the Hispanic community, and other minority
communities, are more likely to walk, bike, and take public transportation. These forms of
transportation put them more at risk of being a pedestrian victim in a vehicular accident. A 2019
study corroborated these findings using a regression analysis that compared the distribution of
cycling infrastructure and various demographic statistics. The study determined that worse
access to bike lanes was most often seen in communities with low educational attainment, high
rates of Hispanic residents, and a lower composite socioeconomic status (Braun, Rodriguez, and
Gordon-Larsen 2019).
Unequal access to cycling infrastructure between different social groups is not a
coincidence, as there is evidence that issues related to equity have often been overlooked when
planning and implementing cycling infrastructure (Cunha and Silva 2022). Cunha and Silva
present a literature review on studies that discuss the distribution of cycling infrastructure within
communities. They determined that overall, cycling infrastructure has often been implemented
more in wealthier and privileged communities than historically underserved communities. This
suggests that change is needed in municipal cycling infrastructure planning processes. It is
critical that the equity impacts of projects are considered.
21
Age is another important demographic statistic that has a relationship with bikeability, as
some age groups may be in more need of safe infrastructure than others. Communities with a
higher population of children have been observed to have higher crash risk (Cottrill and
Thakuriah 2010), which has significant implications for neighborhoods with many children
walking or biking to school. A qualitative study that interviewed 186 parents in Texas gained
insight into the barriers to cycling that children face. It determined that the existence of quality
sidewalks and crosswalks as well as other built-environmental factors played the largest role in
determining whether or not children would walk or bike or drive to school (Kweon et al. 2006).
In communities with a high youth population, extra care must be taken when implementing
cycling infrastructure, as it must be able to protect inexperienced and vulnerable riders.
Additionally, if schools spearhead activities related to teaching bike safety or organizing bike-to-
school days, it is likely that ridership will increase (Staunton and Hubsmith 2003). However,
these policies must be implemented in conjunction with infrastructure improvements.
This thesis incorporates crash data as areas with a high volume of crashes are less
bikeable than areas with fewer crashes (Codina et al. 2022). Following the evaluation of
bikeability using contributing input variables, the results are analyzed with a social lens allowing
for relationships between bikeability and various social demographics to be identified. By
acknowledging the disproportionate amount of pedestrian and vehicular accidents in areas with a
larger youth population, lower median income, and/or higher non-white percentage, city planners
can strive to improve access to safe bike lanes when designing future infrastructure
improvements.
22
2.3 GIS Methodologies and Bikeability
GIS is a powerful tool that can be applied to perform analyses involving spatial data. In
terms of studies related to bikeability and the methodologies used in this thesis, GIS was used to
quantify the bikeability of Redwood City and evaluate the equity impact of proposed cycling
infrastructure projects.
2.3.1 Variables Used to Quantify Bikeability
GIS overlay methodologies have been used to gauge the bikeability of a study area
(Arellana et al. 2020; Codina et al. 2022; Grigore et al. 2019; Grisé and El-Geneidy 2018; Krenn,
Oja, and Titze 2015; Olgun 2020; Porter et al. 2020; Winters et al. 2013). Bikeability
quantification using an overlay methodology is a method of geospatial analysis used in this
thesis. Previous studies that have used similar methodologies each used their own unique set of
input variables that are combined in a weighted sum to quantify bikeability. While each set is
different, there are commonalities between studies in terms of variables used (Table 1).
23
Table 1. Variables used in previous bikeability studies
Variable Studies
Bike lanes
Arellana et al. (2020); Codina et al. (2022); Grigore et al. (2019); Krenn, Oja,
and Titze (2015); Porter et al. (2020); Winters et al. (2013)
Building aesthetics Arellana et al. (2020)
Connectivity Codina et al. (2022); Grisé and El-Geneidy (2018); Winters et al. (2013)
Crime rate Arellana et al. (2020)
Destinations Porter et al. (2020); Olgun (2020); Winters et al. (2013)
Estimated cycling trips Grisé and El-Geneidy (2018)
Existing cycling trips Grisé and El-Geneidy (2018)
Hazards
Arellana et al. (2020); Codina et al. (2022); Grigore et al. (2019); Grisé and
El-Geneidy (2018)
Lighting Arellana et al. (2020)
Ozone level Porter et al. (2020)
Population density Porter et al. (2020)
Police presence Arellana et al. (2020)
Security camera presence Arellana et al. (2020)
Road quality Arellana et al. (2020)
Slope
Arellana et al. (2020); Codina et al. (2022); Grigore et al. (2019); Krenn, Oja,
and Titze (2015); Olgun (2020); Winters et al. (2013)
Speed limit Arellana et al. (2020); Grigore et al. (2019)
Traffic
Arellana et al. (2020); Codina et al. (2022); Grigore et al. (2019); Olgun
(2020)
Tree canopy
Arellana et al. (2020); Grigore et al. (2019); Krenn, Oja, and Titze (2015);
Olgun (2020); Porter et al. (2020)

Among the variables most commonly used in the eight investigated studies are bike lanes,
destinations, hazards, slope, speed limit, and tree canopy. These variables and their utility to
bikeability studies are described in sections 2.3.1.1 through 2.3.1.6.
2.3.1.1 Bike lanes
In bikeability studies, bike lanes as a variable refers to the presence of designated bike
infrastructure within a study area. The existence of official bike lanes alone elevates an area’s
24
bikeability, as it allots cyclists their own space to ride. However, the quality and effectiveness of
the bike lane can vary depending on the type of lane or pavement condition.
Quantifying bike lanes for use in an overlay methodology has been performed in previous
studies. Arellana et al. (2020) applied a polyline overlay methodology where each road segment
was assigned a bikeability score. Each road segment received a bike lanes attribute that
functioned in a binary manner. If a given road had a designated bike lane, it received a score of
one, but if it did not have a designated bike lane, it received a score of 0 (Arellana et al. 2020).
Winters et al. (2013) differed in that a raster weighted overlay was conducted to perform a
bikeability analysis. The presence of bike lanes was still used as an input factor, but in the form
of a bicycle route density raster layer. To create this raster, a feature layer of existing designated
bike lanes was analyzed using the Line Density tool in ArcGIS. The result was a raster
representing the presence of bike lanes that could be integrated into a weighted overlay (Winters
et al. 2013). The use of the Line Density tool also allowed for a non-binary approach, meaning
that raster cells that may not contain their own bike lanes, but are adjacent to bike lanes, receive
a boost in conductivity to bikeability. This is more realistic, as residents would most likely be
willing to travel a few blocks on unprotected paths to reach bikeable infrastructure.
2.3.1.2 Destinations
Destinations is a generalized variable that differs slightly between each study. It relates to
the zoning and land use of a study area, as some zones and uses are more conducive to
bikeability and would induce more residents to cycle. Types of destinations that fit these
characteristics include parks and transit stations (Porter et al. 2020) as well as commercial,
education, entertainment, and office facilities (Winters et al. 2013). Winters et al. (2013)
quantified the location and distribution of destinations by selecting parcels that fit the
25
aforementioned destination types from a broader land use dataset in a GIS. A new layer was
created from the selection featuring only parcel polygons that represented destination locations
and was converted into a point file. The point file was then input into the ArcGIS Point Density
tool to produce a raster showing the density of destination locations across the Vancouver
Metropolitan Area (Winters et al. 2013). This raster was then used in the final bikeability
weighted overlay evaluation.
2.3.1.3 Hazards
Hazards is another generalized variable that may refer to a variety of factors that
influence bikeability. Broadly, hazards refer to potential dangers that a cyclist may encounter
while riding their bike. One example of a hazard is the existence of physical bike lane
obstructions including poles (Arellana et al. 2020) and tram tracks and crossings (Grigore et al.
2019). Hazards may also refer to dangerous intersections along bike routes (Grisé and El-
Geneidy 2018) and the number of cyclist-involved collisions along a given road (Codina et al.
2022). Codina et al. (2022) used hazards as one of their five input variables for their bikeability
evaluation in Barcelona. This was done by dividing the study area into a grid of 100 meter (m)
by 100m cells and then calculating the rate of collisions per bike ride for each cell. This process
resulted in a raster layer that could be used as an input layer into the weighted overlay analysis
(Codina et al. 2022).
2.3.1.4 Slope
Slope has a significant impact on bikeability. If the steepness of a street is over 10
degrees, it is extremely difficult to cycle uphill and would deter most average bikers (Arellana et
al. 2020). Slope is a common variable in bikeability evaluations. Krenn et al. (2015) used slope
as one of five input factors in their bikeability evaluation of Graz, Austria. To evaluate the
26
bikeability in Graz, a weighted overlay of five input rasters was conducted. Each raster was
100m by 100m. In the slope raster, each cell held a value that represented the mean slope of that
cell. These values were later reclassified to fit within a common one to seven scale that could be
used to compare all five rasters together (Krenn et al. 2015). Winters et al. (2013) also
incorporated slope as one of five input variables in their bikeability evaluation of the Vancouver
Metropolitan Area. To quantify slope, a 30m-by-30m digital elevation model (DEM) was
acquired and used as the input for the Slope geoprocessing tool in ArcGIS. The output was set to
percentage rise which assigned each 30m-by-30m cell value of the maximum slope between
itself and bordering cells. This resulting raster was then used as one of five rasters in the
bikeability weighted overlay (Winters et al. 2013).
2.3.1.5 Speed limit
Street speed limits have a significant impact on the amount of safety and attractiveness
that roads offer cyclists. Fast speeds are less safe for cyclists. In fact, cycling guidelines in
Switzerland, the Netherlands, and Denmark state that cyclists should not share space with cars
when the speed limit exceeds 30 to 40 kilometers per hour, which is approximately 20 to 25
miles per hour (MPH) (Grigore et al. 2019). Grigore et al. (2019) and Arellana et al. (2020) both
used speed limit as an input factor in their bikeability overlays. Rather than perform a weighted
overlay using raster data, both studies performed overlay analysis on street segments, with
segment speed limit acting as one input variable. Grigore et al. (2019) acquired speed limit
spatial data for the study area, Basel, Switzerland, from the Office of Mobility Basel-Stadt while
Arellana et al. (2020) acquired data that represented the motorized transport speed of each road
segment in the study area, Barranquilla, Colombia, from the 2012 Master Mobility Plan for
Barranquilla.
27
2.3.1.6 Tree canopy
A strong tree canopy or high percentage of tree coverage can increase the level of
bikeability by improving the aesthetics of the bike ride (Arellana et al. 2020) and cooling down
urban areas during warm periods of the year (Hinterthuer 2019). Arellana et al. (2020) quantified
trees by evaluating each road segment in the study area for the presence of trees. The scoring
system was binary and if a segment had trees, it received a score of one, and if a segment did not
have trees, it received a score of zero (Arellana et al. 2020). Krenn et al. (2015) incorporated
green cover, which includes trees, shrubs, and grasses, into their raster overlay evaluation of
bikeability. For each 100m-by-100m cell within the study area, the total area of green coverage
was calculated. Ultimately, each cell contained a value that quantified its total green coverage
and this raster was used as one of five inputs into the bikeability evaluation overlay (Krenn et al.
2015).
2.3.2 GIS Methodologies to Quantify Bikeability
Using GIS to evaluate the bikeability of a study area is not a novel application of spatial
sciences and multiple studies have been successful at doing so in the past.
A 2020 geospatial research study conducted in Portland, Oregon aimed to quantify
bikeability based on the location of desirable destinations (retail, schools, parks, transit stations,
etc.) and surveyed individual’s biking habits as the primary input information (McNeil 2011).
The study determined that the average Portland cyclist is willing to travel up to 2.5 miles by bike
(roughly 20 minutes) but is willing to travel 22% further along a protected bike lane, which is
physically separated from car lanes, as opposed to a bike lane, which provides no physical
separation from passing cars. This study geocoded a number of locations people would bike to
on a regular basis such as stores, parks, schools, and other amenities. Each amenity type was
28
assigned a score, with the total sum of scores across all amenity types being 100. Thirteen
arbitrary locations in varying neighborhoods and of varying distance from existing bike
infrastructure were then selected as ‘origin points’ and a raster illustrating each point’s service
area, with breaks at 1 mile, 2 miles, and 2.5 miles, was created (see Figure 9).

Figure 9. Portland service area breaks

Next, the bikeability score for each origin point was created by calculating the sum of all
amenity scores in the area, where point deductions were imposed for amenities further from the
origin point. This allowed the researchers to create a map which shows bikeability score at each
origin point, allowing them to deduce trends in bikeability across the city (Figure 10).

29

Figure 10. Portland bikeability scores

Bikeability scores for Portland, OR show a range from 51 (break area 2) to 100 (break
areas 20, 24, 25, and 26) and overall suggest bikeability is higher in the western half of the city
compared to the eastern half of the city. In addition, the study also determined that a lack of road
connectivity, which refers to the density of intersections and ease of moving around a street
network, was the biggest factor weighing down areas with poor bikeability (McNeil 2011).
While it provided valuable insight regarding bikeability, the Portland study did not use a
weighted overlay, which is another useful strategy to assess the overall bikeability of an entire
study area.
In contrast to McNeil’s Portland study, a research study performed across the Vancouver,
Canada Metropolitan Area used a weighted overlay to evaluate bikeability (Winters et al. 2013).
For this study, five input rasters were used: topography, destination density, street connectivity,
30
bike route separation, and bike route density (see Figure 11). The Vancouver project created a
raster output where each pixel holds a value that defines the level of bikeability. The result is a
continuous map that quantifies bikeability in the Vancouver area (Figure 11). Winters’
methodology was precedent for a similar study in Graz, Austria that used a weighted overlay
with cycling infrastructure, presence of separated bicycle pathways, main roads without parallel
bicycle lanes, green and aquatic areas, topography, and land-use as input factors (Krenn, Oja,
and Titze 2015). The study successfully quantified bikeability within the study area. These types
of overlay methodologies and symbology techniques serve as inspirations for the methodology
devised in this thesis.

Figure 11. Input factors and bikeability for the Vancouver Metropolitan Area

31
A third study that implemented a weighted overlay to assess bikeability was undertaken
in Barranquilla, Colombia. This study was unique due to the use of community engagement to
determine which bike infrastructure projects should be prioritized (Arellana et al. 2020). The
project began by designing an overlay methodology to create a bikeability index. Local bikers
were interviewed to provide insights into popular biking origins and destinations, and this
information was overlaid with the bikeability index to determine project priority.
2.3.3 Evaluating the Equity Impact of Cycling Infrastructure
Overlay methodologies can also be used to gauge the extent to which proposed bike lanes
may mitigate issues related to inequity as discussed in section 2.2. By overlaying polygons that
represent certain equity elements, such as income, ethnicity, or state-identified equity zones, it is
possible to deduce which parts of the study area are faced with the most challenges related to
inequality. However, prior to beginning the analytical process, it is important to have background
knowledge regarding previous studies performed in a similar vein, and these are reviewed here.
In addition to assessing bikeability, this thesis analyzed a set of existing proposed bicycle
infrastructure projects in Redwood City and determined which ones should be prioritized in
terms of implementation by the city. While the exact methodology used in this thesis is novel,
previous studies have performed similar analyses for similar reasons. In 2018, an analysis was
carried out in Auburn, Alabama to determine where new bike share stations should be located as
well as in what order their installation should be prioritized. The researchers began by using a
weighted overlay methodology to determine which parts of the city would have the highest
demand for bike share facilities. They then overlaid the resulting map with a map of existing
bike share facilities. Next, they used visual observation, demographic statistics, and an activity-
32
transit accessibility index to determine the impact and utility of the proposed locations to
ultimately decide which sites to prioritize (Jehn, Atiquzzaman, and LaMondia 2018).
Grisé and El-Geneidy (2018) aimed to develop a prioritization ranking for new bike lanes
in Quebec City, Canada. A weighted overlay was generated with input factors that included
existing and expected trips based on survey results, the locations of suggested bike lanes,
dangerous intersections, and network connectivity. This resulted in the identification of locations
where bike lanes should be prioritized. The authors continued their research by considering
equity and historic social inequality within the study area. An index was created to identify
communities based on median household income, unemployment, immigration, and relative
housing cost. These areas roughly overlapped with areas that should be prioritized based on the
weighted overlay results (Grisé and El-Geneidy 2018).
2.4 Urban Design Techniques for Improved Bikeability
There are a variety of effective methods that can be applied to improve the bikeability of
a study area. These methods include public engagement, municipal policy, physical
infrastructure, and the use of geodesign as a guiding methodology.
2.4.1 Public Engagement
A city can boost its urban design process using policy tools and infrastructure upgrades,
and improvements can be maximized by allowing residents and other local stakeholders to have
input during the design process. Engaging citizens through polls, workshops, and meetings
allows decision makers to provide designs that best serve their constituents.
When hosting a workshop with members of the public, it is important to feature
interactive components where participants can easily contribute feedback (Mueller et al. 2018).
This is especially true in design and spatial sciences, where interactive maps and renderings
33
maximize public participation and best engage stakeholders. However, it is critical that the
interactive component is designed in a simple and accessible way, to avoid overcomplicating the
task of public stakeholders and turning them away from contributing feedback (Mahyar et al.
2016). Prior to hosting a public engagement session, it is necessary to be aware of common
pitfalls found when working with stakeholders to best mitigate them. For example, public
engagement workshops have the potential to become argumentative, or foster unfair situations in
which one individual or interest group dominates the conversation (Duan 2021). By
implementing group work, limiting the amount of time people can speak, and using engaging
techniques to boost interest, certain drawbacks can be avoided (Duan 2021).
2.4.2 Policy Tools
Policy tools to improve bikeability can be written based on the results of research and
observation. If crafted well, municipal policy can increase the number of cyclists within a region.
In Copenhagen, Denmark, municipal goals have been set to reduce carbon emissions and
improve general well-being. For example, the city aims to have 50% of work and school
commutes be completed via cycling. Policymakers have developed several tools to assess
progress and move towards completing these goals. The techniques include bi-annual indicators
and metrics and stakeholder and public engagement (Nielsen, Skov-Petersen, and Carstensen.
2013).
Hull and O’Holleran (2014) describe a methodology involving literature and case study
reviews on instances where cities have successfully created cycling-conducive infrastructure. As
a guiding principle, the research claimed that it is imperative that governments consider bicycles
as, at minimum, equal agents when compared to cars. Cities can then adjust their urban plans and
policy to form neighborhoods with a diverse mixture of land use where resources are readily
34
available within a small distance to all modes of transportation. Lastly, the research indicates that
dissuading car use through car taxes and use restrictions can effectively encourage cycling (Hull
and O’Holleran 2014).
2.4.3 Physical Infrastructure
Urban connectivity refers to the extent to which a city is easily navigable and traversable
and is dependent on the street layout and available transit options. Galpern et al. (2018)
monitored cell phone GPS data over a six-year period to assess the mobility of college students
in Calgary, Canada. The researchers determined that students living in neighborhoods with low
street connectivity were more inactive and car-reliant than students living in neighborhoods with
high street connectivity and who were more mobile and reliant on walking and biking. This
suggests that good street connectivity increases the walkability and bikeability of a city (Galpern
et al. 2018). In general, creating urban environments with high population density, a good
mixture of land use, and good walking and biking infrastructure (including high connectivity) are
additional physical infrastructure characteristics that boost bikeability and minimize car
dependence (Saelens, Sallis, and Frank 2003).
Existing literature also outlines the types of cycling infrastructure that are most effective.
While the addition of any cycling infrastructure increases ridership (Parker et al. 2013), the lane
design is still critical to maximize the amount of people who are comfortable using the new
infrastructure. In nearly any context, fully protected bike lanes are the optimal type of
infrastructure. Cycle tracks, and bike boulevards on slow streets, are the only types of bike
infrastructure that demonstrate a notable decrease in crash risk (DiGioia 2017) and reduce injury
risk (Thomas and DeRobertis 2013). However, it is not enough to construct a standalone, fully
protected bike lane. A 2021 study in Sydney, Australia used a predictive transport mode choice
35
model to estimate the number of cyclists in two design scenarios. The first scenario involved the
implementation of a singular cycleway while the second scenario involved the implementation of
an entire network. The model predicted that while the first scenario will increase cycling, it
mainly catered to male, high-income, and older social groups. However, the second scenario
catered to a wider variety of social groups (Standen et al. 2021). This implies that while a city
must start by implementing an initial protected bike lane, they must continue towards a complete
network. The construction of a complete and bikeable network would induce demand, increasing
the number of cyclists. Fosgerau et al. (2023) investigated the relationship between bicycle
infrastructure and induced demand in Copenhagen using a series of spatial simulations. They
determined that the existence of a complete network increases the number of bicycle trips by
59% and the total distance cycled by residents by 88% (Fosgerau et al. 2023).
2.4.4 Geodesign as a Tool to Improve the Built Environment
While urban design has tremendous influence over improving bikeability, geodesign
shows great potential for making cities more livable at the human level by using a mixed-
methods and holistic approach. Geodesign combines geospatial analysis with architectural and
urban design fundamentals as well as community engagement to make decisions regarding
infrastructure and planning, making it a complex, interdisciplinary field (Campagna 2016).
Geodesign manages to combine analytical tools with human creativity (Li and Milburn 2016),
resulting in the creation of enjoyable and useful environments.
36
Chapter 3 Methods
This thesis utilizes an interdisciplinary and mixed methods geodesign approach. Applying a
geodesign methodology allows for the consideration of a multitude of interdisciplinary fields
when making decisions and emphasizes the importance of sustainable design. This thesis makes
use of technical skills found in the fields of GIS, architecture, and urban design and combines
them with soft skills such as public engagement and iterative and sustainable design to advance
bikeability in Redwood City.
3.1 Methods Overview
The methodology applied for this thesis is split into five distinct steps. The first three
steps were all completed individually, before being assessed together in step four which fed into
step five.
First, the existing state of bikeability for each road segment in Redwood City was
assessed in ArcGIS Pro by using a linear weighted sum overlay of spatial data relevant to
bikeability.
Second, a workshop involving Redwood City residents was held to solicit input on the
existing state of cycling within Redwood City as well as residents’ wishes for future
infrastructure improvements. The workshop findings were quantified and added to ArcGIS Pro
for further spatial analysis.
Third, an analysis on the extent to which each proposed project impacted historically
underserved communities was performed in ArcGIS Pro using four datasets that described the
spatial extent of local underserved communities. Each proposed project was assigned an equity
score based on their spatial relationship with the four aforementioned datasets.
37
Fourth, a weighted sum methodology was used in ArcGIS Pro to give each proposed
project a prioritization score. This determined the order that the implementation of proposed
projects should be. The highest scoring projects were located in areas with low bikeability,
underserved populations, and areas that received specific mentions by residents during the
workshop.
Fifth and lastly, design concepts and renderings were created in SketchUp for the three
highest priority projects. An overview of the project methodology can be seen in Figure 12.

Figure 12. Methodology overview

3.2 Bikeability Quantification
Developing a bikeability score for each road segment within Redwood City involved
spatial analysis in ArcGIS Pro using six input variables that impact bikeability. The data
representing each variable were project into the 1983 NAD California Zone III State Plane
coordinate system, which ensures a high level of geographic accuracy for data in Redwood City.
38
The data were then reclassified to a common scale and combined using a weighted sum to
produce a final bikeability score for each street segment. The weights for each input variable
were selected based on weights used for identical variables in similar studies.
3.2.1 Variable Choices and Data Preparation
The six variables used in this study were selected based on data availability as well as the
frequency in which they were used in previous literature. Of the eight investigated bikeability
studies, speed limit was used two times (Arellana et al. 2020; Grigore et al. 2019), destinations
(zoning designation in this thesis) was used three times (Olgun 2020; Porter et al. 2020; Winters
et al. 2013), slope was used six times (Arellana et al. 2020; Codina et al. 2022; Grigore et al.
2019; Krenn, Oja, and Titze 2015; Olgun 2020; Winters et al. 2013), tree canopy was used five
times (Arellana et al. 2020; Grigore et al. 2019; Krenn, Oja, and Titze 2015; Olgun 2020; Porter
et al. 2020), the presence of bike lanes (existing bike lane access in this thesis) was used six
times (Arellana et al. 2020; Codina et al. 2022; Grigore et al. 2019; Krenn, Oja, and Titze 2015;
Porter et al. 2020; Winters et al. 2013), and hazards (crashes in this thesis) was used three times
(Arellana et al. 2020; Grigore et al. 2019; Grisé and El-Geneidy 2018). No other input variable
found in the eight studies was used more than once. These six datasets were acquired and
prepared for use in this thesis and can be seen in Table 2 along with other data that was used in
this project.

39

Table 2. Required data
Data Layer Utility Source Format Date Last Updated
CA Healthy Places
Index
Equity analysis California Healthy Places Index Polygon 2022
C/CAG Equity
Focus Areas
Equity analysis C/CAG Github Polygon Unknown
Census Tracts Create equity polygon data US Census Polygon (Shapefile) 2022
Crashes Bikeability input variable Transportation Injury Mapping System Point 2022
DEM Bike lane access variable USGS Raster (10m) 11 August 2022
Demographic
Block Groups
Equity analysis San Mateo County GIS Open Data Polygon (Shapefile) 17 December 2015
Existing Bike
Lanes
Bike lane access variable RWC WBTI Polyline (KML) June 2022
MTC Equity
Priority Areas
Equity analysis MTC GIS Open Data Polygon (Shapefile) 25 May 2021
Parks Workshop quantification San Mateo County GIS Open Data Polygon (Shapefile) 5 June 2015
Proposed Bike
Lanes
Indicate location of
proposed bike lanes
RWC WBTI Polyline (KML) June 2022
Redwood City
Boundary
Clip input boundary and
feature on maps
California Open Data Portal Polygon (Shapefile) 1 January 2016
Redwood City
Neighborhoods
Cartography Vbeckley_RWC on ArcGIS Online Polygon (ArcGIS Online) 29 January 2018
SamTrans Equity analysis ShockleyD_Samtrans on ArcGIS Online Polygon (ArcGIS Pro) 11 February 2022
Schools Workshop quantification San Mateo County GIS Open Data Polygon (Shapefile) 5 June 2015
Speed Limit Bikeability input variable County of San Mateo GIS Data Download Polyline (Shapefile) 2022
Tree Canopy Bikeability input variable
Multi-Resolution Land Characteristics
Consortium
Raster (30m) 2016
Zoning
Designation
Bikeability input variable Redwood City Community GIS Polygon Unknown
40
3.2.1.1 Speed limit
Speed limit was selected as an input variable for the bikeability evaluation. Traffic speed
has a significant impact on bikeability, as higher speeds are more dangerous for cyclists and
make for a less appealing bike ride. In addition, speed limit was an input factor in some of the
reviewed bikeability studies (Arellana et al. 2020; Grigore et al. 2019).
A polyline dataset was downloaded to ArcGIS Pro from the San Mateo County
Government GIS Data Download website (Information Services 2022) and projected. This
dataset contained road segments for each road within San Mateo County. Each segment features
a variety of attribute information including street type, street name, and speed limit. This dataset
was clipped using the Redwood City boundary dataset, which was downloaded from the
California Government GIS Data Portal (California Open Data Portal 2019). The resulting
polyline feature layer featured every road segment in the study area, including each segment’s
speed limit. This dataset was later modified to include additional fields that account for the other
five bikeability input variables and was therefore renamed BikeabilityVariables.
A new field titled SL_Value was then created to host the reclassified values for speed
limit. To properly conduct a weighted sum of multiple variables, each variable had to maintain
an identical scale. A scale of one to 10 was chosen where 10 indicates highest conduciveness to
bikeability and a score of one indicates lowest conduciveness to bikeability. Each of the six
layers was transformed from a raw value to a numeric value and finally to a scaled value, fit for
use in a weighted sum. Speed limits were reclassified and scaled as per Table 3. The scaled value
for each speed limit was chosen based on existing literature that discusses the relative safety and
comfort of biking on roads with different speed limits. Grigore et al. (2019) stated that the
maximum speed cyclists should be expected to bike with cars is around 20mph. After that, the
41
level of protection offered to cyclists should sequentially increase at approximately 25mph,
40mph, and 45mph (Grigore et al. 2019). As such, speed limits below 20mph were scaled to the
maximum value, 10, while faster speed limits quickly dropped off in scaled value.
Table 3. Speed limit reclassification
Raw
Value
Reclassified
Value
Scaled
Value
0 MPH 0 10
10 MPH 10 10
15 MPH 15 10
25 MPH 25 8
30 MPH 30 6
35 MPH 35 6
45 MPH 45 3
65 MPH 65 1

To reclassify the values, the newly created SL_Value attribute was calculated using the
following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑆𝑃𝐸𝐸𝐷𝐿𝐼𝑀𝐼𝑇
𝑣𝑎𝑟 𝑛𝑥 = 𝑊 ℎ𝑒𝑛 (𝑥 == 0, 10,
𝑥 == 10, 10,
𝑥 == 15, 9, (1)
𝑥 == 25,
7, 𝑥 == 30, 5, 𝑥 == 35, 4, 𝑥 == 45, 3, 𝑥 == 65, 1, ′𝑁𝑂𝑁𝐸 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑥

where x refers to the speed limit attribute of the road segment as set by $feature.SPEEDLIMIT
and nx represents the new speed limit classification of the road feature, dependent on the value of
x. The script instructs ArcGIS Pro to iterate over the features in the road segment dataset and
assigns values to the SL_Value attribute based on the value of the SPEEDLIMIT attribute.
42
3.2.1.2 Zoning designation
Zoning designation serves as a means of identifying locations within Redwood City that
are more likely to correspond with an increase in biking. For example, cycling in and around
mixed-use and destination (parks, schools, retail, dining, etc.) facilities is more preferable to
cycling in and around industrial and single-family areas. Zoning, and its derivative, destinations,
was used as an input factor in some of the reviewed bikeability studies (McNeil 2011; Olgun
2020; Porter et al. 2020; Winters et al. 2013).
The zoning dataset was downloaded from the Redwood City Open Data Portal website
(Redwood City GIS n.d.) in polygon format and added to the ArcGIS Pro project file where it
was projected. A new field was added in the attribute table and was titled GeneralizedZone. This
was done because the Redwood City zoning code has 38 unique designations and for the purpose
of this project, the number of zoning designations was simplified. This is also in alignment with
literature that generalizes zoning to identify destinations of interest (McNeil 2011). The
generalized zoning designations created for this project are industrial, single-family homes,
multi-family homes, destination (commercial, office, parks, schools, etc.), and mixed-use. The
new field was calculated using the following Arcade script:
𝑣𝑎𝑟 𝑧 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑍𝑂𝑁𝐼𝑁𝐺
𝑣𝑎𝑟 𝑔𝑧 = 𝑤 ℎ𝑒𝑛 (𝑧 == ′𝑅𝐻 ′, ′𝑆𝑖𝑛𝑔𝑙𝑒 𝐹𝑎𝑚𝑖𝑙𝑦 ′, 𝑧 == ′𝑅 − 1′, ′𝑆𝑖𝑛𝑔𝑙𝑒 𝐹𝑎𝑚𝑖𝑙𝑦 ′,
𝑧 == 𝑅 ′
− 2
′
, 𝑆 ′
𝑖𝑛𝑔𝑙𝑒 𝐹𝑎𝑚𝑖𝑙 𝑦 ′
, 𝑧 == 𝑅 ′
𝐺 ′
, 𝑆 ′
𝑖𝑛𝑔𝑙𝑒 𝐹𝑎𝑚𝑖𝑙 𝑦 ′
,
𝑧 == 𝑅 ′
− 3
′
, 𝑀 ′
𝑢𝑙𝑡𝑖 𝐹𝑎𝑚𝑖𝑙 𝑦 ′
, 𝑧 == 𝑅 ′
− 4
′
, 𝑀 ′
𝑢𝑙𝑡𝑖 𝐹𝑎𝑚𝑖𝑙 𝑦 ′
,
𝑧 == 𝑅 ′
− 5
′
, 𝑀 ′
𝑢𝑙𝑡𝑖 𝐹𝑎𝑚𝑖𝑙 𝑦 ′
, 𝑧 == 𝑃 ′
𝑂 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
,
𝑧 == 𝐶 ′
𝐴 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
, 𝑧 == 𝐶 ′
𝑁 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
,
𝑧 == 𝐶 ′
𝐵 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
, 𝑧 == 𝐶 ′
𝐺 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
,
𝑧 == 𝐶 ′
𝑃 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
, 𝑧 == 𝐶 ′
𝑂 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
,
𝑧 == 𝐼 ′
𝑅 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
, 𝑧 == 𝐿 ′
𝐼 𝐼 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
, (2)
𝑧 == 𝐼 ′
𝑃 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
, 𝑧 == 𝐺 ′
𝐼 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
, 𝑧
== 𝑇 ′
𝑃 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
, 𝑧 == 𝐴 ′
𝐺 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
,
𝑧 == 𝐼 ′
𝑆 ′
, 𝐼 ′
𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎 𝑙 ′
, 𝑧 == 𝑃 ′
𝐹 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
,
43
𝑧 == 𝑀 ′
𝐻 ′
, 𝑆 ′
𝑖𝑛𝑔𝑙𝑒 𝐹 𝑎 𝑚𝑖𝑙 𝑦 ′
, 𝑧 == 𝐶 ′
𝐵 𝑅 ′
, 𝐷 ′
𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜 𝑛 ′
,
𝑧 == 𝑀 ′
𝑈𝐶 − 𝐸𝐶 𝑅 ′
, 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
, 𝑧 == 𝑀 ′
𝑈𝐶 − 𝑉 𝐵 ′
, 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
,
𝑧 == 𝑀 ′
𝑈𝐶 − 𝑅 𝐶 ′
, 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
, 𝑧 == 𝑀 ′
𝑈𝐶 − 𝑆 𝐵 ′
, 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
,
𝑧 == 𝑀 ′
𝑈𝐶 − 𝐺 𝐵 ′
, 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
, 𝑧 == 𝑀 ′
𝑈 𝑁 ′
, 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
,
𝑧 == ′𝑀𝑈𝑇 ′, ′𝑀𝑖𝑥𝑒𝑑 𝑈𝑠𝑒 ′, 𝑧 == ′𝑀𝑈𝑊 ′, ′𝑀𝑖𝑥𝑒𝑑 𝑈𝑠𝑒 ′, ′𝐶𝑈𝑆𝑇 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑔𝑧 ;

where z refers to the zoning attribute of the zoning polygon feature as set by $feature.ZONING
and gz represents the new generalized zoning classification of the zoning polygon feature,
dependent on the value of z. The script instructs ArcGIS Pro to iterate over the features in zoning
polygon dataset and assigns values to the GeneralizedZone attribute based on the value of the
ZONING attribute.
The zoning data was linked to the roads data using a spatial join to associate each road
segment with a single zone value. In the Spatial Join tool configuration pane, the target feature
was the BikeabilityVariables feature layer. The join feature was the zoning polygon dataset. The
intersect match option was used with a search radius of 100 feet. In the fields section, a new field
in the BikeabilityVariables layer was created titled GenZone and was set equal to the
GeneralizedZone field from the zoning layer. The tool was run, and the result was a single
zoning designation for each road segment. Of the 2,920 road segments, 26 of them had null
values under GenZone. These null values were manually replaced with accurate zoning
information based on visual observation.
The generalized zone attributes were converted from a text format into numeric values to
fit the one to 10 scale. A new field called Zone_Value was created in the BikeabilityVariables
feature layer attribute table and contained numeric values, each of which was associated with a
generalized zoning category as per Table 4. Mixed-use and destination facilities are the most
conducive to bikeability, which is why they were assigned scores of nine and 10 respectively.
44
Residential areas are moderately conducive to bikeability, with multi-family housing being
slightly more conducive to bikeability than single-family housing. Industrial land uses are the
least conducive to bikeability.
Table 4. Zoning designation reclassification
Raw Value
Reclassified
Value
Scaled
Value
Destination 1 9
Industrial 2 1
Single Family 3 4
Mixed Use 4 10
Multi Family 5 6

To reclassify the values, the newly created Zone_Value attribute was calculated using the
following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐺𝑒𝑛𝑍𝑜𝑛𝑒
𝑣𝑎𝑟 𝑛𝑥 = 𝑤 ℎ𝑒𝑛 (𝑥 == ′𝐷𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜𝑛 ′, 9, 𝑥 == ′𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 ′, 1,
𝑥 == 𝑆 ′
𝑖𝑛𝑔𝑙𝑒 𝐹𝑎𝑚𝑖𝑙 𝑦 ′
, 4, (3)
𝑥 == 𝑀 ′
𝑖𝑥𝑒𝑑 𝑈𝑠 𝑒 ′
, 10,
𝑥 == ′𝑀𝑢𝑙𝑡𝑖 𝐹𝑎𝑚𝑖𝑙𝑦 ′, 6, ′𝑁𝑂𝑁𝐸 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑥

where x refers to the generalized zone attribute of the road segment as set by $feature.GenZone
and nx represents the new zoning classification of the road feature, dependent on the value of x.
The script instructs ArcGIS Pro to iterate over the features in the road segment dataset and
assigns values to the Zone_Value attribute based on the value of the GenZone attribute.
3.2.1.3 Slope
Slope was selected as an input variable as the slope of the road or bike lane has a
significant impact on the amount of use the infrastructure can be expected to receive. Biking is
most efficient on a flat surface and if the slope is too steep, it is possible that certain individuals
45
would be unable or unwilling to cycle. Slope was also used as an input factor in some of the
reviewed bikeability studies (Arellana et al. 2020; Codina et al. 2022; Grigore et al. 2019; Krenn,
Oja, and Titze 2015; Olgun 2020; Winters et al. 2013).
A DEM with 10m spatial resolution that encompassed the study area was downloaded
from the United States Geological Survey website (National Map n.d.), added to the ArcGIS Pro
project file, and projected. The DEM was clipped to the Redwood City boundary. The Slope
geoprocessing tool was configured with the DEM as the input to produce the slope layer. The
Slope tool was run and a raster layer covering the study area where each pixel contained a value
representing the mean slope of that pixel in degrees was generated.
To assign each road segment line with a slope value, the Add Surface Information
geoprocessing tool was used. The input feature was the BikeabilityVariables feature layer, and
the input surface was the slope raster. The output property was set to average slope. The tool was
then run and resulted in the creation of a new field within the road segment attribute table. This
field was titled Avg_Slope and featured a numeric value representing the average slope along the
associated road segment.
The slope values were reclassified from a range of zero to 90 degrees to fit within the
scaled value range. A new field was created in the BikeabilityVariables attribute table and titled
Slope_Value. This field hosted the scaled values (see Table 5). Literature has suggested that
a slope of more than five degrees is uncomfortable for cyclists and a slope of more than 10
degrees is nearly impossible for the average biker (Olgun 2020). The scaled values selected for
the slope input variable reflect these conclusions.
46
Table 5. Slope reclassification
Raw Value
(slope in
degrees)
Reclassified
Value
Scaled
Value
10-90 1 1
5-10 5 5
1.5-5 8 8
0-1.5 10 10

To reclassify the values, the newly created Slope_Value attribute was calculated using the
following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐴𝑣𝑔 _𝑆𝑙𝑜𝑝𝑒
𝑣𝑎𝑟 𝑛𝑥 = 𝑤 ℎ𝑒𝑛 (𝑥 < 1.5, 10,
𝑥 ≥ 1.5 && 𝑥 < 5, 8, (4)
𝑥 ≥ 5 && 𝑥 < 10, 5,
𝑥 >= 10, 1, ′𝑁𝑂𝑁𝐸 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑥

where x refers to the slope attribute of the road segment as set by $feature.Avg_Slope and nx
represents the new slope classification of the road feature, dependent on the value of x. The script
instructs ArcGIS Pro to iterate over the features in the road segment dataset and assigns values to
the Slope_Value attribute based on the value of the Avg_Slope attribute.
3.2.1.4 Tree canopy
Tree canopy was selected as an input variable for two reasons. First, a strong tree canopy
implies an abundance of shade, which can help keep cyclists cool when they are biking during
high temperatures. Second, exposure to trees, and nature in general, boosts the aesthetic of biking
and makes it a more enjoyable experience. Tree canopy was used as an input variable in some of
the reviewed bikeability studies (Arellana et al. 2020; Grigore et al. 2019; Krenn, Oja, and Titze
2015; Olgun 2020; Porter et al. 2020).
47
The NLCD 2016 US Forest Service Tree Canopy Cover was downloaded from the Multi-
Resolution Land Characteristics Consortium website (MRLC Data n.d.) at a spatial resolution of
30m and added to the ArcGIS Pro project file before being projected and clipped. Each pixel in
the raster contained a single value that represents the percentage of tree canopy coverage within
that pixel.
To assign each road segment line with a tree cover value, the Add Surface Information
geoprocessing tool was used. The input feature was the BikeabilityVariables feature layer and the
input surface was the tree cover raster. The output property was set to “Mean Z”, which in this
instance is average tree coverage percentage of each pixel. This resulted in the creation of a new
field within the road segment attribute table. This field was titled Z_Mean and featured a numeric
value representing the average tree coverage percentage along the associated road segment. The
field name was immediately changed to TreeCanopy to avoid later confusion.
The tree canopy values were reclassified from a range of one to 10 to fit within the scaled
value range. A new field was created in the BikeabilityVariables feature layer attribute table and
titled TC_Value. This field hosted the scaled values, which are visible in Table 6.
Table 6. Tree canopy reclassification
Raw Value (Tree
Canopy Percentage)
Reclassified
Value
Scaled
Value
0-10% 1 3
10-20% 2 4
20-30% 3 5
30-40% 4 6
40-50% 5 7
50-60% 6 8

48
To reclassify the values, the newly created TC_Value attribute was calculated using the
following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑇𝑟𝑒𝑒𝐶𝑎𝑛𝑜𝑝𝑦
𝑣𝑎𝑟 𝑛𝑥 = 𝑊 ℎ𝑒𝑛 (𝑥 < 10, 3, 𝑥 ≥ 10 && 𝑥 < 20, 4,
𝑥 ≥ 20 && 𝑥 < 30, 5, 𝑋 ≥ 30 && 𝑥 < 40, 6, (5)
𝑥 ≥ 40 && 𝑥 < 50, 7, 𝑥 ≥ 50 && 𝑥 < 60, 8,
𝑥 ≥ 60, 9, ′𝑁𝑂𝑁𝐸 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑥

where x refers to the tree canopy attribute of the road segment as set by $feature.TreeCanopy and
nx represents the new tree canopy classification of the road feature, dependent on the value of x.
The script instructs ArcGIS Pro to iterate over the features in the road segment dataset and
assigns values to the TC_Value attribute based on the value of the TreeCanopy attribute.
3.2.1.5 Bike lane access
The existence of and ease of access to bike lanes is an important variable in quantifying
bikeability. If bike lanes are sparce and difficult to access, ridership will inevitably be low. Bike
lane access was used as an input variable in some of the reviewed bikeability studies (Arellana et
al. 2020; Codina et al. 2022; Grigore et al. 2019; Krenn, Oja, and Titze 2015; Porter et al. 2020;
Winters et al. 2013).
Developing data to represent bike lane access required more analytical and preparatory
work than the other five variables. The existing and proposed bike lane data were retrieved from
the Redwood City WBTI website as a KML file. Upon addition to the ArcGIS Pro project file,
the layer was input into the Feature Class to Feature Class geoprocessing tool to allow for
modifications and the output layer was projected. The downloaded data did not feature explicit
attributes that distinguished between class or between whether a bike lane was existing or
proposed. Instead, each lane was assigned an ID value that could be used to identify the class and
49
status. However, to improve legibility of the layer, two new fields were added: Lane_Class (I, II,
III, or IV) and Status (existing or proposed). The lane class field was calculated using the
following Arcade script:
𝑣𝑎𝑟 𝑖𝑑 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑆𝑦𝑚𝑏𝑜𝑙𝐼𝐷
𝑣𝑎𝑟 𝑐𝑙𝑎𝑠𝑠 = 𝑤 ℎ𝑒𝑛 (𝑖𝑑 == 0, ′𝐶𝑙𝑎𝑠𝑠 𝐼 ′,
𝑖𝑑 == 1, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 𝐼 ′
,
𝑖𝑑 == 2, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼𝐼 𝐼 ′
,
𝑖𝑑 == 3, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 𝑉 ′
, (6)
𝑖𝑑 == 4, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 ′
,
𝑖𝑑 == 5, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 𝐼 ′
,
𝑖𝑑 == 6, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼𝐼 𝐼 ′
,
𝑖𝑑 == 7, ′𝐶𝑙𝑎𝑠𝑠 𝐼𝑉 ′, ′𝑎 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑐𝑙𝑎𝑠𝑠 ;

where id refers to the bike lane class and status attribute of the bike lane feature as set by
$feature.SymbolID and class represents the new bike lane class value of the bike lane feature,
dependent on the value of id. The script instructs ArcGIS Pro to iterate over the features in the
bike lane dataset and assigns values to the Lane_Class attribute based on the value of the
SymbolID attribute.

The lane status field was calculated using the following Arcade script:
𝑣𝑎𝑟 𝑖𝑑 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑆𝑦𝑚𝑏𝑜𝑙𝐼𝐷
𝑣𝑎𝑟 𝑐𝑙𝑎𝑠𝑠 = 𝑤 ℎ𝑒𝑛 (𝑖𝑑 == 0, ′2022′ ,
𝑖𝑑 == 1, 2
′
022
′
,
𝑖𝑑 == 2, 2
′
022
′
,
𝑖𝑑 == 3, 2
′
022
′
, (7)
𝑖𝑑 == 4, 𝑃 ′
𝑟𝑜𝑝𝑜𝑠𝑒 𝑑 ′
,
𝑖𝑑 == 5, 𝑃 ′
𝑟𝑜𝑝𝑜𝑠𝑒 𝑑 ′
,
𝑖𝑑 == 6, 𝑃 ′
𝑟𝑜𝑝𝑜𝑠𝑒 𝑑 ′
,
𝑖𝑑 == 7, ′𝑃𝑟𝑜𝑝𝑜𝑠𝑒𝑑 ′, ′𝑎 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑐𝑙𝑎𝑠 𝑠 ;

50
where id refers to the bike lane class and status attribute of the bike lane feature as set by
$feature.SymbolID and class represents the new bike lane status value of the bike lane feature,
dependent on the value of id. The script instructs ArcGIS Pro to iterate over the features in the
bike lane dataset and assigns values to the Status attribute based on the value of the SymbolID
attribute.
The bike lane access attribute information describes whether a road segment is within
200m of a bike lane or not. 200m was chosen based on similar scenarios in the investigated
literature (Winters et al. 2013). Cyclists may be dissuaded to bike if they are required to travel
significantly further to get to safe infrastructure. There are four possible values for this attribute:
Class I/IV, Class II, Class III, and none. Classes I and IV were combined as they both provide
physical separation from the road for cyclists, the only difference being that Class I bike lanes do
not follow existing roads while Class IV bike lanes are adjacent to existing roads. If a road
segment is within 200m of multiple lane classes, the value would be determined by the highest-
ranking class (Classes I and IV are the best, followed by Class II and then Class III). The bike
lane access calculations were performed using two separate ModelBuilder tools.
To determine which street segments were within 200m of a bike lane, a ModelBuilder
tool (Figure 13) was created and ran four times, once for each bike lane class. First, the Polyline
to Raster geoprocessing tool was used to convert a road segment polyline layer into a raster
layer. The cell size was set to 1 foot to minimize the effects of using a raster representation of
streets rather than a polyline representation. The Path Distance tool was then used. The input
feature data was the bike lane layer (Class I, II, III, or IV) and the input cost raster was the output
of the Polyline to Raster geoprocessing tool. The maximum distance was set equal to 656.168
feet, as this is equal to 200m. The result of the Path Distance tool was a new raster layer where
51
each cell contained a value that described how far it was from the bike lanes. Cells from the
original street segment raster layer that were further than 200m from the bike lane were
eliminated from the Path Distance tool output. The Reclassify tool was then used to convert each
cell value from its respective distance value to a value of one to score all cells within 200m of a
bike lane equally. The reclassified raster layer was then used as the input raster in the Raster to
Polygon geoprocessing tool. The polygonal output layer was then used in the Select Layer by
Location geoprocessing tool, where the input feature was the BikeabilityVariables (shown as
SJ_Zoning in Figure 13) layer and the relationship was set to intersect. The layer with selection
was then input into the Copy Features geoprocessing tool, which created a new line dataset
containing all road segments within 200m of the specified bike lane class. The ModelBuilder was
run a total of four times, with the only alteration for each run being the bike lane layer input in
the Path Distance tool. The end result was four new line layers: road segments within 200m of
Class I bike lanes, road segments within 200m of Class II bike lanes, road segments within 200m
of Class III bike lanes, and road segments within 200m of Class IV bike lanes. The Class I layer
and Class IV layer were merged together as both refer to similar infrastructure types with full
physical protection for cyclists from cars.
52

Figure 13. Bike lane access ModelBuilder layout

Next, the four separate line layers were combined into a single layer where each road
segment was assigned a lane class, or “none,” dependent on its spatial relationship with existing
cycling infrastructure. This was done using a ModelBuilder tool (Figure 14). The Erase
geoprocessing tool was used to erase Class II road segments from Class III road segments. The
resulting output, titled Class III Erase, was then used as the input feature in another Erase
geoprocessing tool, where the erase feature was the Class I/IV road segment layer.
Simultaneously, the Class I/IV layer was erased from the Class II layer, resulting in an output
titled Class II Erase. Class III Erase, Class II Erase and the Class I/IV layer were all merged,
53
resulting in a street segment layer where each segment that was within 200m of a bike lane had a
lane class value listed under the newly created field, BLAClass. The merged layer was then
erased from a layer of all street segments within the study area. This resulted in a layer of street
segments that were over 200m from any class of bike lane. This layer was then merged with the
bike lane merge layer. A spatial join was then conducted to attach the BLAClass attribute to the
BikeabilityVariables feature layer. While street segments within 200m of a bike lane had a value
for the BLAClass attribute, those that were further than 200m had a null value. This was changed
to a value of “None” using the following Arcade script in the Calculate Field tool:
𝑣𝑎𝑟 𝑏 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐵𝐿𝐴𝐶𝑙𝑎𝑠𝑠
𝑣𝑎𝑟 𝑛𝑏 = 𝑤 ℎ𝑒𝑛 (𝑏 == ′𝐶𝑙𝑎𝑠𝑠 𝐼 /𝐼𝑉 ′, ′𝐶𝑙𝑎𝑠𝑠 𝐼 /𝐼𝑉 ′,
𝑏 == 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 𝐼 ′
, 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 𝐼 ′
, (8)
𝑏 == ′𝐶𝑙𝑎𝑠𝑠 𝐼𝐼𝐼 ′, ′𝐶𝑙𝑎𝑠𝑠 𝐼𝐼𝐼 ′, ′𝑁𝑜𝑛𝑒 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑏

where b refers to the bike lane class attribute of the road segment as set by $feature.BLAClass,
and nb represents the new bike lane class value of the bike lane, dependent on the value of b. The
script instructs ArcGIS Pro to iterate over the features in the road segment dataset and if a road
segment has a null bike lane class attribute value, it is assigned “None”.
54

Figure 14. Bike lane access merge ModelBuilder layout

At this point, the bike lane access attribute was in text form and had to be reclassified to
fit within the scaled value range. A new field was created in the BikeabilityVariables feature
layer attribute table and titled BLA_Value. This field hosted the scaled values, which are visible
in Table 7. The scaled values take into account the level of comfort of using each of the four
bike lane access scores. Class I and IV bike lanes are ideal and were scaled to a value of nine.
Class II bike lanes are moderately attractive options and were scaled to a value of 5. Class III
bike lanes and road segments that were further than 200m of any bike lane were scaled to low
values, as these road segments are not very conducive to high bikeability.
55
Table 7. Bike lane access reclassification
Raw Value (Bike
Lane Access)
Reclassified
Value
Scaled
Value
200m of C1 & C4 Class I/IV 9
200m of C2 Class II 5
200m of C3 Class III 3
>200m from any None 1

To reclassify the values, the newly created BLA_Value attribute was calculated using the
following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐵𝐿𝐴𝐶𝑙𝑎𝑠𝑠
𝑣𝑎𝑟 𝑛𝑥 = 𝑊 ℎ𝑒𝑛 (𝑥 == ′𝐶𝑙𝑎𝑠𝑠 𝐼 /𝐼𝑉 ′, 9,
𝑥 == 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼 𝐼 ′
, 5, (9)
𝑥 == 𝐶 ′
𝑙𝑎𝑠𝑠 𝐼𝐼 𝐼 ′
, 3,
𝑥 == ′𝑁𝑜𝑛𝑒 ′, 1, ′𝑁𝑂𝑁𝐸 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑥

where x refers to the bike lane class attribute of the road segment as set by $feature.BLAClass
and nx represents the new bike lane access classification of the road feature, dependent on the
value of x. The script instructs ArcGIS Pro to iterate over the features in the road segment dataset
and assigns values to the BLA_Value attribute based on the value of the BLAClass attribute.
3.2.1.6 Crashes
Crash data serves as a means of representing the hazards of biking along a given street
and to a lesser extent, the expected amount of traffic on a given street. Streets with a high amount
of crashes are more dangerous and may incite more interactions between cyclists and drivers.
While crashes were not explicitly used as an input in the reviewed studies, hazards and traffic
counts were (Arellana et al. 2020; Codina et al. 2022; Grigore et al. 2019; Grisé and El-Geneidy
2018), both of which can be partially represented by crash statistics.
56
A point dataset was downloaded from the University of California, Berkeley
Transportation Injury Mapping System website (Transportation Injury Mapping System n.d.) that
contains point data of traffic accidents. The data was filtered so that only crash points within San
Mateo County between March 2015 and March 2020 where a cyclist was involved were
downloaded. This temporal scale was chosen as it was still relatively recent, but also protected
against reduced traffic and crash numbers as a result of the COVID-19 pandemic. The point data
was added to the ArcGIS Pro project file and was projected. The data was then clipped using the
Redwood City boundary polygon layer.
Next, a relationship between the crash point data and the road line segment data had to be
created. A 40-foot buffer was created around each road segment using the Buffer geoprocessing
tool. This ensured that all points along roadways would be included and that in the case of a
crash at an intersection, both intersecting roads would be considered as the location of a crash.
The Aggregate Points tool was used to sum the number of crash points within each road segment
buffer polygon. A Join function with the completely within join operation was used to join the
sum of crashes within each buffer with the BikeabilityVariables feature layer. This was done in a
new field called Crashes where each road segment’s attribute was a value representing the
number of cyclist-involved crashes along the road segment between March 2015 and March
2020. Next, the data was reclassified to the scaled values listed in Table 8. Road segments with
zero crashes were scaled to a value of nine, because they have historically been safe streets for
cyclists. However, a value of 10 was not given since the chance of a crash always exists. Road
segments with five crashes were scaled to a value of three because while these segments
represent what have historically been the most dangerous roads and intersections for cyclists
57
over the past five years, one crash per year is not an alarming rate. As such, five-crash-segments
were not assigned a value of one.
Table 8. Crashes reclassification
Raw Value (Number
of Crashes)
Reclassified
Value
Scaled
Value
0 0 9
1 1 7
2 2 6
3 3 5
4 4 4
5 5 3

A new field was created in the BikeabilityVariables feature layer attribute table and titled
Crash_Value. This field was calculated using the following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐶𝑟𝑎𝑠 ℎ𝑒𝑠
𝑣𝑎𝑟 𝑛𝑥 = 𝑊 ℎ𝑒𝑛 (𝑥 == 0, 9,
𝑥 == 1, 7,
𝑥 == 2, 6, (10)
𝑥 == 3, 5,
𝑥 == 4, 4,
𝑥 == 5, 3, ′𝑁𝑂𝑁𝐸 ′)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑛𝑥

where x refers to the crash attribute of the road segment as set by $feature.Crashes and nx
represents the new crash classification of the road feature, dependent on the value of x. The script
instructs ArcGIS Pro to iterate over the features in the road segment dataset and assigns values to
the Crash_Value attribute based on the value of the Crashes attribute.
58
3.2.2 Weighted Sum to Quantify Bikeability
To quantify bikeability, a weighted sum equation was used. The equation was applied to
each individual road segment and combined the six input variables with their respective weights
to output a single number that represented the bikeability of that particular road segment.
A new text field was created on the BikeabilityVariables feature layer and titled
Bikeability. This field was calculated using a weighted sum methodology within the Calculate
Field tool. Each of the six input variables was assigned a weight as per Table 9. The weights
for this project roughly follow weights used in previous literature (Arellana et al. 2020; Codina et
al. 2022; Grigore et al. 2019; Grisé and El-Geneidy 2018; Krenn, Oja, and Titze 2015; Porter et
al. 2020; Winters et al. 2013). For example, many previous studies allotted higher weights to
existing bike lane access and slope as opposed to other input variables. Certain variables have
more of an impact on bikeability than others and that is reflected in the weights selected for this
analysis.
Table 9. Bikeability weights
Variable Weight
Speed Limit 15%
Zone 15%
Slope 20%
Tree Canopy 8%
Bike Lane
Access
30%
Crashes 12%
Total 100%

59
For each road segment, each reclassified variable score was multiplied by its weight.
Each of these six multiplications were then added together, resulting in a final bikeability score
within the range of one to 10. The calculation was performed using the following Arcade script:
𝑣𝑎𝑟 𝑏𝑖𝑘𝑒𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = (𝑉𝑎𝑙𝑆𝑝𝑒𝑒𝑑𝐿𝑖𝑚𝑖𝑡 ∗ 0.15)
+ (𝑉𝑎𝑙𝐺𝑒𝑛𝑍𝑜𝑛𝑒 ∗ 0.15)
+ (𝑉𝑎𝑙𝑆𝑙𝑜𝑝𝑒 ∗ 0.20)
+ (𝑉𝑎𝑙𝑇𝑟𝑒𝑒𝐶𝑎𝑛𝑜𝑝𝑦 ∗ 0.08) (11)
+ (𝑉𝑎𝑙𝐵𝐿𝐴𝐶𝑙𝑎𝑠𝑠 ∗ 0.30)
+ (𝑉𝑎𝑙𝐶𝑟𝑎𝑠 ℎ𝑒𝑠 ∗ 0.12)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑏𝑖𝑘𝑒𝑎𝑏𝑖𝑙𝑖𝑡𝑦

where bikeability represents the bikeability score of a road segment, ValSpeedLimit represents
the speed limit attribute of a road segment, ValGenZone represents the generalized zone attribute
of a road segment, ValSlope represents the slope attribute of a road segment, ValTreeCanopy
represents the tree canopy attribute of a road segment, ValBLAClass represents the bike lane
access attribute of a road segment, and ValCrashes represents the crash attribute of a road
segment. The numeric values which are used to multiply each respective attribute represent the
overall weight of the associated attribute as it pertains to evaluating bikeability.
3.2.3 Reclassifying Bikeability for Site Selection
Bikeability scores were ultimately used in the final site selection process, which is
outlined in section 3.5. To comply with the other site selection input variables which are outlined
in sections 3.33.3.3 and 3.4, the bikeability scores were reclassified to a range of zero to four.
First, a new field was created within the proposed project polyline feature layer attribute
table and was titled Bikeability. Because bikeability scores had previously been created for each
street segment in Redwood City, it was only necessary to transfer the values into the proposed
projects attribute table. However, because the proposed projects do not perfectly align with the
60
street segments, it was necessary to use a ModelBuilder geoprocessing tool (Figure 15) to assign
each proposed project with a bikeability score.

Figure 15. Proposed project bikeability ModelBuilder layout

The proposed bike lane layer was set as the input layer for the Buffer geoprocessing tool.
A buffer distance of 50ft was used to ensure that proposed projects would not pick up on
bikeability scores from parallel streets while also ensuring that small differences in the two
polylines could be accounted for. The output buffer was then used as the input polygon in the
Summarize Within tool. The dataset to be summarized was the bikeability attribute of the
BikeabilityVariables feature layer (named Overlay_v2 in Figure 15). The Summarize Within tool
calculated the mean bikeability score of all road segments within each respective buffer polygon.
The output was a polygon layer with an identical geographic extent as the buffer layer, but a new
attribute of mean bikeability. A spatial join was then performed where the bikeability value of
each polygon was joined to the proposed projects polyline feature layer using the completely
within match option. The output of this was a polyline layer featuring each proposed project
61
including a bikeability attribute value. This attribute value could then be joined to a new attribute
field titled “MeanBikeability” in the proposed project feature layer using a common attribute,
such as project name.
To execute the weighted sum function, it is necessary for each of the input variables to
maintain an identical scale. Because the equity sum attribute value had a range of zero to four
due to the use of four datasets of equity-related polygons, it was necessary to reclassify the
bikeability values to be within this range. Furthermore, the bikeability values were inverted
because in the context of project site selection, a lower bikeability score is more desirable;
proposed projects that are in locations with poor bikeability as of 2022 should be prioritized over
projects in locations with high bikeability. This reclassification was completed by first adding a
new field titled BikeReclass and then calculating that field using the following Arcade script:
𝑣𝑎𝑟 𝐵𝑖𝑘𝑒𝑅𝑒𝑐𝑙𝑎𝑠𝑠 = 4
− (($𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑀𝑒𝑎𝑛𝐵𝑖𝑘𝑒𝑎𝑏𝑖𝑙𝑖𝑡𝑦 ∗ 4) (12)
/ 10)
𝑟𝑒𝑡𝑢𝑟𝑛 𝐵𝑖𝑘𝑒𝑅𝑒𝑐𝑙𝑎𝑠𝑠

where BikeReclass is equal to the new reclassified bikeability attribute and
$feature.MeanBikeability refers to the bikeability attribute of the proposed project layer.
3.2.4 Demographic Comparison
To gain insight into the relationship between bikeability and social demographic
information in Redwood City, a methodology using light geospatial analysis and symbolization
techniques was used.
After the bikeability layer was created, a demographic polygon layer was downloaded
from the San Mateo County GIS open data portal (San Mateo County GIS n.d.). These layers
were aggregated at the block group level and included a variety of attributes, including the
62
percentage of households that did not own a car and median household income. First, the new
layer was added to the ArcGIS Pro project file where it was projected and clipped to the
Redwood City boundary polygon layer. The Spatial Join geoprocessing tool was used to add the
average bikeability within each block group to the demographic attribute table. Bivariate
symbology was then applied to the layer to visualize the relationship between bikeability and no-
car households, bikeability and median household income, and bikeability and non-White
percentage. A total of six maps were produced with three of them illustrating the demographic
information in Redwood Shores and three of them illustrating the demographic information in
the rest of Redwood City.
3.3 Workshop
A workshop was organized to solicit input from Redwood City residents on the current
and future state of bikeability in the city. The workshop required the formation of a relationship
with local government agencies and community stakeholders. Details on planning, execution,
and post-workshop activities are described below.
3.3.1 Workshop Planning
To prove effective, the workshop had to be methodically planned out. It was critical that
the materials of the workshop supported high-quality discussions and participation. Furthermore,
planning was necessary to ensure that participants would attend the meeting.
Following initial conversations with Redwood City Mayor Giselle Hale, the workshop
planning was initiated with Rafael Avendaño at the nonprofit organization Redwood City
Together. This nonprofit works closely with the city and looks to improve issues of inequity in
the city. One of their focuses is on improving bike safety, especially for children. Mayor Hale
also connected the author with City Manager Melissa Stevenson Diaz, who proceeded to loop in
63
Transportation Manager Jessica Manzi. Ms. Manzi then brought WBTI Project Manager Malahat
Owrang into the discussion. Ms. Owrang provided existing and proposed bike infrastructure
spatial data.
Meanwhile, an initial email was sent to Mr. Avendaño of Redwood City Together to
describe the thesis project and express desire to work together to plan a mutually beneficial
workshop. Mr. Avendaño was receptive to the idea and set up a 45-minute meeting between the
author, Mr. Avendaño, and Jackie Campos, who is the nonprofit’s Safe Routes to School
specialist.
In the meeting, Mr. Avendaño and Ms. Campos spoke about Redwood City Together and
the organization’s goals and methods. They also spoke about the ongoing Safe Routes to School
project and why it is important that the city’s youth have easy and efficient access to schools and
parks via foot or bike. The author then discussed the project and geodesign methodology before
outlining the importance of a workshop and explaining how it would be a mutually beneficial
partnership. The author would be able to provide meaningful public feedback as well as GIS
analysis while Redwood City Together and the City of Redwood City would help with
participant outreach and planning. Ultimately, Mr. Avendaño and Ms. Campos were both
interested in helping the author host a workshop, but it would first require confirmation from
Redwood City Together Leadership Council members. The members would be attending the
upcoming Safe Routes to School task force meeting, so Ms. Campos blocked out time in the
agenda for the author to present the workshop idea and conduct a vote on whether or not the
workshop would be confirmed.
At the Safe Routes to School task force meeting, there were approximately 20 attendees.
The author was able to present the workshop proposal and Ms. Campos proceeded to poll the
64
attendees about whether or not they approved of the proposal. All but one approved, which
meant that the workshop was permitted to move forward. In the coming days, Ms. Campos and
the author held a separate meeting with the one member who disapproved of the proposal to
further discuss the workshop’s purpose and methods. The member ultimately approved of the
idea.
At this point, the workshop had been approved. Because of the nature of the workshop
and the fact that the author would be interacting with human subjects and incorporating their
responses into the project, Institutional Review Board (IRB) approval was deemed necessary.
The author filled out the required paperwork to seek IRB approval through the University of
Southern California (USC). The author also completed a required CITI Program module on
performing a Social-Behavioral study on Human Subjects. The application was submitted for
review. About one month later, it was determined that due to the nature of the workshop and the
fact that all results would remain anonymous indefinitely, IRB approval was not needed and the
workshop was ethically allowed to take place
For the next couple of months, the author met with Ms. Campos to advance the planning
process of the workshop. This involved brainstorming the format, determining a date, and
preparing methods of participant outreach.
Ultimately, it was determined that the workshop would take place on February 23, 2023
over Zoom. The author originally planned to have an in-person workshop, but the decision was
made to host it online in part due to the ongoing COVID-19 pandemic. Outreach and the
development of final workshop materials began in early February. Ms. Campos spearheaded the
outreach and contacted several Redwood City Together connections as well as local schools and
cyclist groups. She also created a flier with workshop information that was distributed to mailing
65
lists (Figure 16) as well as an Eventbrite site where interested individuals could RSVP and
receive a Zoom link.

Figure 16. Workshop outreach flier.
66

The author also engaged in the outreach process, speaking at the February Safe Routes to
School task force meeting and sharing a brief slideshow with information about the workshop.
Participants of this meeting were invited to the workshop.
The author created content to be used in the workshop. The first deliverable was a
slideshow, created in Google Slides. The slideshow began with a brief introduction of the author,
a workshop agenda, and housekeeping items related to maintaining anonymity in participant
responses. Next, the discussion shifted to cover the topic of bikeability and why it is critical that
cities adopt improved cycling infrastructure. The presentation then transitioned to a more
focused discussion on cycling in Redwood City, showing images of the four classes of bike lanes
as well as maps of bikeability as of 2022. The introductory presentation ended by talking about
problems and solutions facing bikeability in Redwood City as well as outlining the work the city
has already done through the WBTI. The next couple of slides included links to the survey and
Miro boards, which served as the interactive components of the workshop. Following the
activities, the presentation would resume, first by allowing the present members of the Redwood
City Team of the Silicon Valley Bicycle Coalition to show a slide and speak about their group.
The author then completed the workshop by displaying two concluding slides that summarized
the workshop and offered some final thoughts on bikeability.
The Google Form was created and incorporated questions that were designed to elicit
responses about the participants’ experiences cycling in Redwood City as well as what they
would like to see in the future in terms of bike infrastructure. There was a total of eight
questions, which are listed in Table 10.
67
Table 10. Workshop survey questions
Question Answer Format
How important is living in a bikeable city to you?
(Select one)
Not important, minimally important,
neutral, important, very important
Do you ride a bike? If so, how often? Free response
If you cycle, what are your reasons for doing so?
(Select all that apply)
I don’t bike, work or school,
errands, exercise, pleasure, to get
outdoors, other (free response)
What do you see as barriers to biking more in
Redwood City? (unprotected lanes, fast drivers,
inconvenient, lack of storage facilities/bike parking
etc.)
Free response
What are some variables besides bike infrastructure
that would make you more likely to cycle? (More
trees, more mixed-use zoning, etc.)
Free response
Select all infrastructure types you would like to see in
Redwood City. (Select all that apply)
Images of: physical lane separation,
bollard lane separation, integrated
roundabouts, conventional bike
lanes, other (free response)
If more bike infrastructure was built, which classes of
lane would you be willing to use? (Select all that
apply)
Class I, Class II, Class III, Class IV,
none
Do you have any other comments to make about
bikeability in Redwood City?
Free response

Miro was selected as the interface to allow participants to engage in an interactive
mapping activity. Miro is an online application which serves as a virtual whiteboard, allowing
users to add text, images, drawings, and shapes. A user can create multiple documents, also
known as Miro boards, and share these boards with other users who can then simultaneously
collaborate in the virtual space. Three activities were created, each on their own Miro board. The
activities were meant to foster collaboration, so three copies of each activity were created on
each board and each board would be used by a group of participants who would be placed in a
breakout room.
68
The first activity allowed participants to draw locations on a map of Redwood City that
they considered dangerous to pedestrians and cyclists (Figure 17). On each Miro board, there
was space below each interactive map where participants could use the sticky note tool to leave
comments explaining their decisions.

Figure 17. Activity 1 Miro board

The second activity allowed participants to draw out their dream cycling network,
indicating on the map where they would like to see improved infrastructure (Figure 18).
69

Figure 18. Activity 2 Miro board

The third activity asked participants to select their top three proposed projects and
indicate which they would most like to see be implemented (Figure 19).
70

Figure 19. Activity 3 Miro board

3.3.2 Conducting the Workshop
The workshop took place from 5:00pm to 6:30pm on Thursday, February 23, 2023. It
included a slide deck presentation, survey, and interactive and collaborative mapping exercise.
71
Twenty participants RSVP’d for the workshop, yet only 11 attended. The workshop went
smoothly. It started right at 5pm and began with the presentation outlined in section 3.3.1. Next,
participants were asked to complete the Google Survey, of which the link was provided in the
Zoom chat. Participants were given 10 minutes to complete the survey and were instructed that
because this may be more time than necessary, extra time could be used as a bathroom, food, or
water break. Upon completion of the survey, the interactive activities began. The link to the first
Miro board was provided and participants were split into breakout rooms. Because two
participants left the workshop early, the nine remaining participants were split into three groups
of three participants. The groups were instructed to draw on their own board and delegate one
drawer to minimize chaos on the Miro board. Groups were given 10 minutes to discuss their
responses to the prompt and draw out their conclusions. After the 10 minutes was up, the author
facilitated a 10-minute discussion where participants could explain their decisions and drawings
and also comment on other group’s conclusions. This process was repeated two more times for
the final two Miro boards.
Following the activities, the presentation was resumed, and the Redwood City Team of
the Silicon Valley Bicycle Coalition provided some information about their organization. The
author then made concluding remarks and thanked the participants for their attendance. The
workshop ultimately ran five minutes over the allotted time, ending at 6:35pm.
3.3.3 Assessment of Workshop Results
The goal of the workshop was to obtain public comments on the current and future state
of cycling in Redwood City as well as determine which of the city’s 132 proposed bike
infrastructure projects were most popular. The content created by workshop attendees was
72
synthesized and organized for the purposes of selecting a cohort of proposals to move forward in
this project’s geodesign workflow.
The author reviewed the Zoom recording and created a Google Sheet document to list
each street or proposed project that was mentioned either in the survey, on the Miro board, or in
verbal discussion. Each street or proposed project that was mentioned was listed on the sheet in a
column. There were 16 total streets and two proposed projects mentioned. Of the 16 mentioned
streets, two of them did not have any associated proposed projects and they were discarded. An
additional column was filled out with the proposed project labels (project number and name)
input next to each associated street. In some cases, a mentioned street had multiple proposed
projects along it and new proposed project rows were added to accommodate this. After
extracting the proposed projects from the mentioned streets and listing them along with the two
directly mentioned in the workshop, there was a total of 24 proposed projects referred to either
directly or indirectly during the workshop (Table 11). Next, a column of values representing the
number of mentions for each street or proposed project was added. Another column was added to
indicate the number of positive mentions for each street or proposed project. Of the 24 proposed
projects, only two did not have 100 percent positive mentions.
73
Table 11. Proposed projects mentioned during the workshop
Proposed Projects Mentioned
34 – Alameda de las Pulgas
36 – Alameda de las Pulgas
98 – Alameda de las Pulgas
119 – Arguello St.
113 – El Camino Real
82 – Hudson St.
64 – Jefferson Ave.
54 – Madison Ave.
17 - Maple St.
20 – Maple St.
122 – Maple St.
37 – Middlefield Rd.
49 – Middlefield Rd.
111 – Middlefield Rd.
120 – Middlefield Rd.
75 – Myrtle St.
18 – Path from Seaport Blvd. to Veterans Blvd. under
U.S. 101
12 – Path through Red Morton Park
72 – Poplar Ave.
59 – Redwood Ave.
60 – Roosevelt Ave.
73 – Vera Ave.
74 – Vera Ave.
121 – Woodside Rd.

In addition to naming specific streets that could benefit from improved cycling
infrastructure, the participants generally agreed on two main considerations that should be kept
in mind when planning out Redwood City’s future bike network. First, when possible, arterial
74
cycling routes should be created to serve as safe routes across the entirety of the city as well as to
maximize the efficiency of bike routes and avoid irregular patterns in the network. Second,
implementing cycling infrastructure near schools and parks should be prioritized over other
locations to increase bike safety and accessibility for the most vulnerable cyclists in the city.
Participants also voiced that this would be a good way to inspire the next generation of Redwood
City residents to adopt cycling as a main form of transportation.
To accommodate the public’s desire for new cycling infrastructure being implemented
near parks and schools, a new attribute field titled ParksSchools was added to the proposed
projects dataset. A dataset downloaded from the San Mateo County GIS data hub website
(Information Services n.d.) that featured polygons of all the schools and parks in San Mateo
County was added to the ArcGIS Pro project file. A 500ft buffer was then added to this layer
using the Buffer geoprocessing tool. The Select Within tool was used to select all proposed
projects that were at least partially within the buffer layer. All of the selected proposed projects
were assigned a value of one in the data table, while those outside of the buffer zone were
assigned a value of zero. The Google Sheet column delineating whether a street or proposed
project was near a school or park could then be filled out by cross checking each proposed
project with the new attribute field values. In the Google Sheet, proposed projects near schools
and parks were given a value of one while proposed projects not near schools and parks were
given a value of zero.
At this point, a final score that quantified the needs of the residents could be computed.
This was done for each of the 24 proposed projects and was done in a manner to ensure that all
scores would be within the range of zero and four. Each mention of the street or proposed project
after the initial mention would add 0.25 points to the final score. Next, the ratio of positive
75
comments was calculated, and the resulting decimal was added to the score. Lastly, one point
was added for proposed projects that were near schools and parks. Once final scores were
tabulated, they were added into the proposed projects layer in ArcGIS Pro. A new field titled
CommEngage was created and calculated using the following Arcade script:
𝑣𝑎𝑟 𝑥 = $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐿𝑎𝑏𝑒𝑙
𝑣𝑎𝑟 𝐶𝑜𝑚𝑚𝐸𝑛𝑔𝑎𝑔𝑒 = 𝑊 ℎ𝑒𝑛 (𝑥 == ′121 − 𝑊𝑜𝑜𝑑𝑠𝑖𝑑𝑒 𝑅𝑑 . ′, 2.17,
𝑥 == 1
′
13 − 𝐸𝑙 𝐶𝑎𝑚𝑖𝑛𝑜 𝑅𝑒𝑎 𝑙 ′
, 2.6,
𝑥 == 6
′
0 − 𝑅𝑜𝑜𝑠𝑒𝑣𝑒𝑙𝑡 𝐴𝑣𝑒 .
′
, 2.25,
𝑥 == 5
′
9 − 𝑅𝑒𝑑𝑤𝑜𝑜𝑑 𝐴𝑣𝑒 .
′
, 2.25,
𝑥 == 6
′
4 − 𝐽𝑒𝑓𝑓𝑒𝑟𝑠𝑜𝑛 𝐴𝑣𝑒 .
′
, 2.25,
𝑥 == 7
′
3 − 𝑉𝑒𝑟𝑎 𝐴𝑣𝑒 .
′
, 2.5,
𝑥 == 7
′
4 − 𝑉𝑒𝑟𝑎 𝐴𝑣𝑒 .
′
, 2.5,
𝑥 == 1
′
7 − 𝑀𝑎𝑝𝑙𝑒 𝑆𝑡 .
′
, 2,
𝑥 == 2
′
0 − 𝑀𝑎𝑝𝑙𝑒 𝑆𝑡 .
′
, 1,
𝑥 == 1
′
22 − 𝑀𝑎𝑝𝑙𝑒 𝑆𝑡 .
′
, 2,
𝑥 == 8
′
2 − 𝐻𝑢𝑑𝑠𝑜𝑛 𝑆𝑡 .
′
, 2,
𝑥 == 3
′
4 − 𝐴𝑙𝑎𝑚𝑒𝑑𝑎 𝑑𝑒 𝑙𝑎𝑠 𝑃𝑢𝑙𝑔𝑎 𝑠 ′
, 2.25, (13)
𝑥 == 3
′
6 − 𝐴𝑙𝑎𝑚𝑒𝑑𝑎 𝑑𝑒 𝑙𝑎𝑠 𝑃𝑢𝑙𝑔𝑎 𝑠 ′
, 1.75,
𝑥 == 9
′
8 − 𝐴𝑙𝑎𝑚𝑒𝑑𝑎 𝑑𝑒 𝑙𝑎𝑠 𝑃𝑢𝑙𝑔𝑎 𝑠 ′
, 1.25,
𝑥 == 5
′
4 − 𝑀𝑎𝑑𝑖𝑠𝑜𝑛 𝐴𝑣𝑒 .
′
, 1,
𝑥 == 1
′
19 − 𝐴𝑟𝑔𝑢𝑒𝑙𝑙𝑜 𝑆 𝑡 .
′
, 2,
𝑥 == 1
′
2 − 𝑃𝑎𝑡 ℎ 𝑡 ℎ𝑟𝑜𝑢𝑔 ℎ 𝑅𝑒𝑑 𝑀𝑜𝑟𝑡𝑜𝑛 𝑃𝑎𝑟 𝑘 ′
, 2.25,
𝑥 == 1
′
8 − 𝑃𝑎𝑡 ℎ 𝑓𝑟𝑜𝑚 𝑆𝑒𝑎𝑝𝑜𝑟𝑡 𝐵𝑙𝑣𝑑 . 𝑡𝑜 𝑉𝑒𝑡𝑒𝑟𝑎𝑛𝑠 𝐵𝑙𝑣𝑑 . 𝑢𝑛𝑑𝑒𝑟 𝑈 . 𝑆 . 101
′
, 1,
𝑥 == 7
′
2 − 𝑃𝑜𝑝𝑙𝑎𝑟 𝐴𝑣𝑒 .
′
, 2,
𝑥 == 7
′
5 − 𝑀𝑦𝑟𝑡𝑙𝑒 𝑆 𝑡 .
′
, 2,
𝑥 == 3
′
7 − 𝑀𝑖𝑑𝑑𝑙𝑒𝑓𝑖𝑒𝑙𝑑 𝑅𝑑 .
′
, 2.25,
𝑥 == 4
′
9 − 𝑀𝑖𝑑𝑑𝑙𝑒𝑓𝑖𝑒𝑙𝑑 𝑅𝑑 .
′
, 1.25,
𝑥 == 1
′
11 − 𝑀𝑖𝑑𝑑𝑙𝑒𝑓𝑖𝑒𝑙𝑑 𝑅𝑑 .
′
, 2.25,
𝑥 == ′120 − 𝑀𝑖𝑑𝑑𝑙𝑒𝑓𝑖𝑒𝑙𝑑 𝑅𝑑 . ′, 2.25, 0)
𝑟𝑒𝑡𝑢𝑟𝑛 𝐶𝑜𝑚𝑚𝐸𝑛𝑔𝑎𝑔𝑒

where x is set equal to $feature.Label which represents the label attribute of the proposed project
and CommEngage represents the community engagement score attribute. The script instructs
76
ArcGIS Pro to iterate over the features in the proposed project dataset and assigns values to the
CommEngage attribute based on the value of the Label attribute.
Another post-workshop step was sending a follow up email to all participants. This email
thanked them for their input and included a link for them to view the workshop recording at their
leisure.
3.4 Equity Analysis
Because funding for infrastructure improvement projects is not infinite, it is important to
strategically prioritize projects for actual implementation. Areas that have historically been
underserved by infrastructure and educational opportunities and that have a relatively low
median household income are often targeted for improvements by cities and developers in an
effort to level the societal playing field. In some instances, infrastructure improvement projects
undertaken in these areas may be eligible for external funding from private, regional, state,
and/or federal sources.
After corresponding with Ms. Owrang, it was decided that the equity quantifications of
four different organizations would be referred to when determining which proposed projects to
prioritize based on their equity impact. Each of the organizations’ websites identify areas within
their respective jurisdictions that have been historically underserved, have a lower-than-average
median household income, or have a significant minority population. The four organizations are
listed below, along with maps of the areas in and around Redwood City that they deem as having
previously been inequitably served by government programs.
The first organization was the City/County Association of Governments (C/CAG) of San
Mateo County. C/CAG used a custom criterion which included three variables to determine
where their equity focus areas were located. These criteria were median household income,
77
race/ethnicity, and the housing and transportation affordability index. Figure 20 shows a map of
the C/CAG equity focus areas in and around Redwood City. The data was acquired from the
C/CAG online map (C/CAG of San Mateo County n.d.). While the data could not be directly
downloaded, it is aggregated at the Census Block Group level. To visualize the C/CAG equity
focus areas in ArcGIS Pro, corresponding block groups were selected from a Census Block
Group layer and a new layer was created from the selection.

Figure 20. C/CAG Equity Focus Areas

The second organization was SamTrans, the transportation agency that runs bus services
within San Mateo County. SamTrans has established their own set of priority areas that have
78
historically been underserved by public transportation. SamTrans actively works to improve bus
route connectivity in these areas. SamTrans determined their equity priority areas based on three
criteria: median household income, the presence of racial and ethnic minorities, and the zero-car
household rate. Figure 21 shows a map of the SamTrans equity focus areas in and around
Redwood City. The data was acquired from ArcGIS Online (ShockleyD_Samtrans 2022). The
link was provided by Ms. Owrang.

Figure 21. SamTrans Equity Planning Areas

The third organization was the Metropolitan Transportation Commission (MTC) which is
the agency that oversees transportation across the nine counties that make up the San Francisco
Bay Area. The MTC had identified census tracts within their jurisdiction that are considered
79
equity priority communities (EPCs). EPCs were determined by looking at the following
demographic criteria: people of color, low income, limited English proficiency, zero-vehicle
households, seniors over 75 years old, people with disabilities, single-parent families, and rent
burdened households. Projects within MTC EPCs are eligible for grants and are therefore more
likely to be implemented as a result of available funding. Figure 22 shows a map of the MTC
EPCs. The data was acquired from the MTC’s open data portal website (MTC GIS 2022).

Figure 22. MTC EPCs

The fourth equity quantification was the California Healthy Places Index (CAHPI). The
CAHPI dataset was created by the Public Health Alliance of Southern California and assigns
neighborhoods across California with a score that quantifies the relative health of a community
80
based on 25 factors, including healthcare access, housing availability, and education quality.
While the CAHPI is not a direct stakeholder, rather it is a data-derived index, Ms. Owrang
recommended it be used to assess the equity impact of the proposed projects. Redwood City
communities below the 50th percentile across the state were selected as being relevant to the
equity analysis. Figure 23 shows a map of the neighborhoods in and around Redwood City that
scored beneath the 50th percentile in the CAHPI. The data was available on the CAHPI website
(Public Health Alliance of Southern California n.d.). While the data could not be directly
downloaded, it is aggregated at the Census Tract level. To visualize the CAHPI Census Tracts in
the bottom 50th percentile in ArcGIS Pro, corresponding tracts were selected from a Census
Tract layer and a new layer was created from the selection.
81

Figure 23. CAHPI polygons

Four new fields were added to the 132 proposed projects’ attribute table, and each field
was named for one of the four equity polygon datasets. A ModelBuilder was developed to assign
each proposed project a value for each of the four attributes (Figure 24). The model used the
Select Layer by Location tool to select proposed projects that overlapped each respective equity
polygon dataset. The selected proposed projects were then used as the input for the Calculate
Field tool. This tool used a short Python script to assign a binary attribute value for each
proposed project. If a proposed project overlapped an equity polygon, the attribute was assigned
a value of one. If a proposed project did not overlap an equity polygon, the attribute was
assigned a value of zero.
82

Figure 24. Equity score ModelBuilder layout

Next, an additional field was created and titled EquitySum. Because there were four sets
of equity polygons, the maximum value for this field was four (if a proposed project overlapped
all four equity polygons) and the minimum value for this field was zero (if a proposed project
overlapped none of the four equity polygons). This attribute was calculated by summing the
binary scores assigned to each individual equity attribute calculated in the ModelBuilder. This
was done by using the following Arcade script within the Calculate Field tool:
𝑣𝑎𝑟 𝐸𝑞𝑢𝑖𝑡𝑦𝑆𝑢𝑚 = ($𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑆𝑎𝑚𝑇𝑟𝑎𝑛𝑠
+ $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝑀𝑇𝐴𝐸𝑃𝐶
+ $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐶𝐴𝐻𝑃𝐼 (14)
+ $𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐶𝐶𝐴𝐺 _𝐸𝐹𝐴 )
𝑟𝑒𝑡𝑢𝑟𝑛 𝐸𝑞𝑢𝑖𝑡𝑦𝑆𝑢𝑚

where EquitySum refers to the total equity score of the proposed project, $feature.SamTrans
refers to the SamTrans attribute value of the proposed project feature layer, $feature.MTCEPCs
refers to the MTC EPC attribute value of the proposed project feature layer, $feature.CAHPI
refers to the CAHPI attribute value of the proposed project feature layer, and
83
$feature.CCAG_EFAs refers to the CCAG Equity Focus Areas attribute value of the proposed
project feature layer.
3.5 Site Selection
This section discusses the methodology applied to prioritize proposed projects and
determine which three would move on to the design stage of the project. This was done by using
a weighted sum equation to compute each proposed project’s prioritization score based on its
level of bikeability, equity impact, and the extent, if any, it received public support during the
workshop.
The final step in calculating the site selection rankings was to perform a weighted sum
combining the proposed project’s ability to improve equity, bikeability, and meet public needs.
Each of the three inputs was assigned a nearly identical weight, with equity being weighted 33%,
bikeability 33%, and public needs 34%. These weights were selected as each of the three input
variables are of relative equal importance. This would also help break ties in the event of
multiple proposed projects scoring the same for a given variable. A new field titled EqBikePe
was added and calculated using the following Arcade script:
𝑣𝑎𝑟 𝑥 = ($𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐸𝑞𝑢𝑖𝑡𝑦𝑆𝑢𝑚 ∗ .33)
+ ($𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐵𝑖𝑘𝑒𝑅𝑒𝑐𝑙𝑎𝑠𝑠 ∗ .33) (15)
+ ($𝑓𝑒𝑎𝑡𝑢𝑟𝑒 . 𝐶𝑜𝑚𝑚𝐸𝑛𝑔𝑎𝑔𝑒 ∗ .34)
𝑟𝑒𝑡𝑢𝑟𝑛 𝑥

where x represents the priority score of a proposed project, $feature.EquitySum represents the
equity score of a proposed project, $feature.BikeReclass represents the bikeability score of a
proposed project, and $feature.CommEngage represents the workshop score of a proposed
project. The numeric values which are used to multiply each respective attribute represent the
overall weight of the associated attribute as it pertains to evaluating bikeability.
84
Once calculated, the EqBikePe attribute could be sorted in descending order in ArcGIS
Pro to identify which projects scored the highest. A map indicating the priority score of each
proposed project was also created.
3.6 Proposed Project Design Concepts
Diagrams and models of infrastructure proposals were created for the three highest-
scoring projects that did not already undergo a city design process. The development of final
design proposals is linear and was repeated for each of the three projects. Google Maps and
spatial data were used to investigate the surrounding contexts of each site, with the investigation
results influencing the final design concept. SketchUp was used to create the final designs. All
designs include fully protected Class I bike lanes, as workshop participants were adamant that
these should be strongly prioritized over other class options.
3.6.1 Site Investigation
The physical and social characteristics of the area surrounding each selected project were
investigated in order to become familiar with the site context. This was done to allow each
project to be optimized for its particular location within Redwood City.
3.6.1.1 Current status assessment
While “walking” down the street using Street View in Google Maps, specific
observations regarding the size of the street, the presence of sidewalk buffers, and the type of
housing were documented. The purpose of this step was to become familiar with the site and
how someone interacting with it may view it. The author was also able to apply their own
experience of living in Redwood City and was able to assess how busy the street may be at any
given time, and whether it should be considered an arterial transit corridor or a street for local
85
traffic. This is important to know because if the road is not heavily trafficked, and there are
parallel alternatives near it, it is possible to consider a road diet. A road diet refers to downsizing
the number of lanes on a road in favor of bike or pedestrian infrastructure or the implementation
of one-way traffic. Screenshots were taken of the Google Street View imagery for each site and
were referenced during the design process.
Another important task to undertake during the status assessment was to locate nearby
schools, parks, retail, and mixed-use facilities. These places serve as key destinations for cyclists
and their location can alter the orientation of the new bike infrastructure design. For example, if
there is a park and retail center on the north side of one of the streets under investigation, it may
be better to have a two-way bike lane on the north side of the street, rather than have one-way
bike lanes on either side of the street. Looking at existing bike infrastructure is also important,
especially if it intersects with the street under investigation, as this can influence the orientation
of the new design. Lastly, it is important to locate existing driveways, as in the new designs,
space will need to be left for cars to enter off-street properties. A map was created in ArcGIS Pro
to visualize nearby schools, parks, mixed-use, retail, and existing bike lane facilities making use
of data downloaded from the San Mateo County and Redwood City GIS data portals and was
referenced during the modeling process.
3.6.1.2 Demographics of the surrounding neighborhood
The demographic investigation focused on the percentage of households that did not own
a car. This is a key metric, because if this number is high, that implies less parking is needed and
the decision to either maintain the existing parking situation or remove parts of it in favor of
added bike or pedestrian infrastructure can be made. Median household income and non-white
86
percentage were two other demographics that were investigated. This information was
documented and referred to during the design modeling phase of the thesis.
3.6.1.3 Zoning designations of the surrounding neighborhood
If there are a lot of single-family parcels, street parking may be more necessary due to
limited garage sizes and the fact that many families own multiple cars. However, if there are a lot
of multi-family parcels, there may be fewer cars on the street due to the existence of apartment
complex garages. Furthermore, dense housing near frequent and reliable transit options may
require less parking as residents could be more inclined to use bikes, SamTrans, or CalTrain as a
primary form of transit. A map of the zoning designations for the parcels surrounding the
proposed project was created in ArcGIS Pro using the original zone data file that was
downloaded from Redwood City’s GIS data portal (Redwood City GIS n.d.) and was used to
inform the modeling process.
3.6.2 Modeling
StreetMix is an online application that allows users to design street cross-sections by
dragging and dropping various street infrastructure components on an online canvas (StreetMix
n.d.). StreetMix was used to create an illustration of the potential cross-section of the newly
designed street. The width of the street was determined by measuring the distance between
property lines across the street in Google Maps. The StreetMix street interface was then set to the
same distance and components such as driving lanes, parking lanes, bike lanes, sidewalks, and
sidewalk buffers were dragged on until the space was filled. The width of each component could
be adjusted depending on size constraints; however, each component was restricted by a
minimum width as determined by StreetMix restrictions as well as local planning regulations.
Once a satisfactory design had been created, a screenshot of the interface was taken.
87
In ArcGIS Pro, the street containing the proposed project of interest was selected from
the polyline layer featuring all of Redwood City’s streets. A new layer was created from the
selection, isolating the proposed project-of-interest street line. The line was then exported as a
.dwg computer-aided design (CAD) file and stored on the desktop.
A new Sketchup Pro file was created. The .dwg file containing the linework of the
proposed project street was imported into the file. The entirety of the line was selected and the
Weld Edges tool was applied, which merged each of the individual street segments into one
continuous line geometry.
The SketchUp to OpenStreetMap (skp2osm) plugin was then downloaded from the
OpenStreetMap (OSM) Wiki. This plugin is compatible with SketchUp and allows the user to
add geometry from OSM directly into a SketchUp file.
Next, OSM was accessed through an internet browser window. The street segment of
interest was centered on the map page. The export button was selected (Figure 25, upper left),
followed by the “manually select a different area” option (Figure 25, upper right). A rectangle
appeared on the page and the author centered it around the road of interest, ensuring that adjacent
buildings, intersections, and roads were all within the rectangle frame. The export button was
then clicked and an .osm file was downloaded, containing all building footprint and road
geometry (Figure 25, bottom).
88

Figure 25. OSM data download steps

Back in the SketchUp file, the .osm file containing the building footprint and road
geometry was imported. The road geometry representing the road of interest was then deleted.
The line previously created by merging the street segments was then selected and moved into the
correct geographic location within the matrix of OSM buildings. Because the OSM selector tool
is a rectangle that is centered on the road of interest, there are building footprints and streets
present in the OSM file that are irrelevant to the study area. These irrelevant features serve no
purpose and only slow down SketchUp processing speeds. The unneeded features were selected
and deleted from the project file.
89
Next, the remaining buildings which are relevant to the proposed project, meaning they
are either adjacent or otherwise close to the street, were given a 3D appearance. This was done
by manually applying the Extrude tool to each building footprint and dragging the footprint
upwards to a certain height. The height was estimated by using Google Street View and Google
Maps in a 3D view.
Next, a cross section of the proposed project was created perpendicular to the end of the
street linework. The cross section of the proposed infrastructure was centered on the endpoint of
the street linework. The cross section must be two-dimensional and cannot be represented by a
line, so it was created one foot deep. Lines were created perpendicular to the cross section and
were located based on the proposed street design. A line parallel to the initial cross section line
was created one foot back, resulting in rectangular areas that represent the various sections of the
street (sidewalk, sidewalk buffer, bike lane, bike lane buffer, parking lane, road lane, etc.).
Each rectangular area was then assigned a material. Sidewalks and bike lane curb barriers
were assigned concrete material. The road and street parking rectangles were assigned asphalt
material. The sidewalk barriers were assigned grass material. The bike lanes were assigned a
green material that was similar in color to the green paint already used by Redwood City to mark
bike lanes.
By this time, a cross section of the proposed street design had been created. It was one
foot long and as wide as the street (for example, 18m) and split up into rectangular sections, each
indicating a different use of the right of way. Lastly, each rectangular section was extruded one
foot downwards, ensuring that a 3D geometry was created with the top maintaining the same
elevation as the road linework and the bottoms of the buildings.
90
Next, the Follow Me tool was used to stretch the cross section along the length of the
street line. Each face of each 3D section (sidewalk, sidewalk buffer, bike lane, etc.) had to be
created individually using the Follow Me tool. First, the Follow Me tool was selected from the
SketchUp toolbar. Next, the line representing the street was selected. Next, the face of the 3D
road segment was selected. The end of the street line was then selected, and the 3D road segment
followed the linework to the opposite end. This was repeated for each 3D component making up
the proposed design. By the end of this process, the proposed design had been stretched to match
the length of the street segment while also maintaining any curves in the street segment line
shape.
Next, each design was customized to fit in with the surrounding built context. The first
step of customizing the street design with the existing surrounding context was adding cross-
streets and creating intersections. This was done using the spaces left behind by roads that were
found in the original OSM map. The center point of the off-shooting road was determined and its
location along the newly designed road was clicked. A line was created between this point along
the edge of the newly designed road and the end of the intersecting road. This line was nine
meters long, as it was meant to be half of the intersecting road width, which was in all cases 18m
wide. This was repeated in the other direction. Once the overall space of the intersecting road
was created, sidewalks and the street could be created using the Line and Material tools. Next,
all sidewalks, sidewalk buffers, and bike lane curbs were extruded six inches upwards, matching
the height of an average curb and resulting in a three-dimensional model.
Various features had to be customized for each site. Using the Line tool in SketchUp
driveways that connected to the street, cutting through bike lanes, were drawn out and assigned
the concrete material. The same was done for walkways that led from the sidewalk to the
91
building entrances. 3D tree models were also imported from the SketchUp 3D warehouse and
placed on both private properties as well as the sidewalk medians. Bike lane stencils were added
to the bike lanes. White squares were also added to the bike lanes at instances where they were
intersected by roads or driveways. The addition of these squares was meant to increase bike lane
visibility and were inspired by similar techniques used in The Netherlands (Figure 26).
Additionally, speed limit signs, stop signs, crosswalk signs, and one-way signs from the
SketchUp 3D warehouse were added as needed.

Figure 26. Netherlands bike lane driveway infrastructure

Once the design was complete, the SketchUp view was changed from a parallel
projection to a perspective view. This allowed the Position Camera and Look Around tools to be
used to position the model in a manner that would be suitable for final rendering. For each
model, one view was created from the perspective of a cyclist or pedestrian and one view was
created from an aerial perspective.
92
V-Ray for SketchUp was downloaded from Chaos and was used to create site renderings.
V-Ray works well with SketchUp and was first installed and added as an extension. It was then
opened, and the settings were adjusted to produce a quality rendering. The lighting setting was
set to dome light, and a sky-blue image was used to provide background coloring. In the general
V-Ray settings, progressive render was turned off, the quality was set to high, the denoiser was
toggled on, the image width was set to 1,920 pixels, image height was set to 1,080 pixels, and a
file path was created to automatically save the output rendering as a .png file. V-Ray was then
run, and renderings were produced. Based on the rendering results, the author could choose to
adjust the camera angle and re-run V-Ray.
Following the design modeling process, six deliverables were created for each of the
three selected proposed projects: an imagery map of the area around the proposed project, an
amenity map of the area around the proposed project, a zoning map of the area around the
proposed project, a StreetMix diagram, and two renderings of the new design.
93
Chapter 4 Results
The objective of the work described in this thesis was to provide conclusions for advancing
Redwood City’s bicycle infrastructure based on a geodesign workflow. The approach was built
around five research elements: (1) quantification of Redwood City’s bikeability as of 2022; (2)
solicitation of input from local residents on the current (2023) and proposed cycling
infrastructure of Redwood City; (3) evaluation of Redwood City’s proposed cycling
infrastructure projects to positively impact issues of economic, social, and transit inequality; (4)
selection of three proposed cycling infrastructure improvement projects that should be most
prioritized; (5) conceptual designs for potential projects that were assessed to be of potential
highest impact.
4.1 Bikeability Quantification
The result of the bikeability assessment for Redwood City including the Redwood Shores
area is depicted in Figure 27. The assessment is the combined output from the analysis of the
posted speed limit, zoning designation, terrain, tree canopy, bike lane access and historic cyclist-
involved accident data. Sections 4.1.1 and 4.1.2 describe the results for Redwood City and
Redwood Shores respectively.
94

Figure 27. Redwood City bikeability
4.1.1 Redwood City Bikeability
In Redwood City, bikeability is overwhelmingly moderate, leaning towards the lower end
of the bikeability spectrum (Figure 28). However, there are clusters of high bikeability within the
city.
95

Figure 28. Redwood City bikeability (excluding Redwood Shores)

There are a few instances of high bikeability in Redwood City, mainly concentrated
around the downtown area in neighborhoods 12, 13 and 14 (see Figure 29). Neighborhoods nine,
10, and 11 are residential, however they support moderate to high bikeability (see Figure 29).
Neighborhoods one, two, three, four, five, six, seven, eight, 15, and 16 support low to moderate
bikeability (see Figure 29). These neighborhoods are almost entirely residential, leaving a
significant amount of the population with no access to bikeable infrastructure. This indicates that
there is significant opportunity for improvement to Redwood City’s cycling network.
96

Figure 29. Redwood City bikeability and neighborhoods

4.1.2 Redwood Shores Bikeability
In Redwood Shores, bikeability is generally high (Figure 30). While there are a few
instances of low bikeability, these street segments are found within quiet suburban areas and are
not a major roadblock to cyclists. Redwood Shores is flat and was master planned with the
inclusion of circumferential bike lanes that allow for straightforward travel throughout the
surrounding area, fully separated from automobile roads. This makes it easy for residents to
travel within Redwood Shores and comfortably ride from their homes to other homes, offices, or
retail centers. The street segments scoring between low and moderate bikeability are generally
97
restricted to single family residential areas where traffic is relatively slow. Despite the generally
high bikeability, the bike infrastructure connecting Redwood Shores with the Bay Trail which
leads into the main Redwood City area is not very developed and suggests that the bikeability
between Redwood Shores and the rest of Redwood City is poor. Overall, Redwood Shores
demonstrates high bikeability and can serve as an inspiration for future urban developments.

Figure 30. Redwood Shores bikeability

4.1.3 Proposed Projects ’ Bikeability
Bikeability was one of three input variables used to determine which proposed projects
would be selected to receive design proposals, with the other two being workshop results and
98
equity impact. The relative bikeability of each proposed project as of 2023 was determined using
a methodology outlined in section 3.2.3. The results of the analysis are shown in Figure 31.
Proposed project bikeability.

Figure 31. Proposed project bikeability scores

4.2 Workshop Findings
The workshop was structured to allow participants to give their input on the current
(2023) and future state of bikeability in Redwood City through various communicative methods.
Eleven total participants attended the workshop, with nine of them staying for the entirety of the
99
event. The nine Redwood City residents that attended the whole workshop all participated in the
interactive activities described below.
The different media for data collection, the Google Survey, Miro board activities, and
verbal discussion sessions, resulted in similar outputs. Participants mentioned several streets and
proposed projects where they would like to see better cycling infrastructure. After reviewing the
survey results, Miro board, and workshop recording, a list of streets that received mentions was
created along with the number of mentions and the number of proposed projects associated with
each street. A map of the mentioned and associated proposed projects was also created (Figure
32).

Figure 32. Proposed projects mentioned in the workshop
100

Workshop participants also emphasized their interest in new cycling infrastructure being
prioritized near parks and schools to create a safer rider environment for young bikers. As part of
the quantification of workshop comments, a map was created to visualize the locations of parks
and schools in Redwood City. A 500ft buffer was also added to identify specific parts of the city
that should be targeted with improved infrastructure (Figure 33).

Figure 33. 500ft buffer around parks and schools

Using the list of mentioned streets, it was possible to extract the proposed projects
associated with each street. In some cases, there was only one proposed project for a given street
101
but in other cases, there were multiple proposed projects along a given street. A table of all 24
proposed projects that were either directly (mentioned in activity 3) or indirectly (associated with
a mentioned street) mentioned during the workshop was created. The table included the proposed
project number and name, the number of mentions, the number of positive mentions, the
percentage of positive mentions, a binary value that indicates whether the project was within 500
feet of a school or park (yes is equal to one and no is equal to zero), and a final public feedback
score. Final public feedback scores are listed in Table 12 for each proposed project that was
mentioned during the workshop.
102
Table 12. Public feedback scores for proposed projects
Proposed Project ID
Number and Name
Mentions
Positive
Mentions
Mention
Score
Percent
Positive
Mentions
Score
Near
School/Park
Score
Public
Feedback
Score
34 - Alameda de las
Pulgas
2 2 0.25 1 1 2.25
36 - Alameda de las
Pulgas
3 3 0.75 1 0 1.75
98 - Alameda de las
Pulgas
2 2 0.25 1 0 1.25
119 - Arguello St. 1 1 0 1 1 2
113 - El Camino
Real
5 3 1 0.6 1 2.6
82 - Hudson St. 1 1 0 1 1 2
64 - Jefferson Ave. 2 2 0.25 1 1 2.25
54 - Madison Ave. 1 1 0 1 0 1
17 - Maple St. 1 1 0 1 1 2
20 - Maple St. 1 1 0 1 0 1
122 - Maple St. 1 1 0 1 1 2
37 - Middlefield Rd. 2 2 0.25 1 1 2.25
49 - Middlefield Rd. 2 2 0.25 1 0 1.25
111 - Middlefield
Rd.
2 2 0.25 1 1 2.25
120 - Middlefield
Rd.
2 2 0.25 1 1 2.25
75 - Myrtle St. 1 1 0 1 1 2
18 - Path from
Seaport Blvd. to
Veterans Blvd. under
U.S. 101
1 1 0 1 0 1
12 - Path through
Red Morton Park
2 2 0.25 1 1 2.25
72 - Poplar Ave. 1 1 0 1 1 2
59 - Redwood Ave. 2 2 0.25 1 1 2.25
60 - Roosevelt Ave. 2 2 0.25 1 1 2.25
73 - Vera Ave. 3 3 0.5 1 1 2.5
74 - Vera Ave. 3 3 0.5 1 1 2.5
121 - Woodside Rd. 3 2 0.5 0.67 1 2.17

Based on the calculation of the public feedback score, El Camino Real ranked the highest
with a score of 2.6 and Vera Avenue ranked the second highest with a score of 2.5. One of the
103
Maple Street projects, the Madison Avenue project, and the U.S. 101 underpass between Seaport
Boulevard and Veterans Boulevard all ranked the lowest with a score of one. Figure 34 shows a
map of all proposed projects and how they scored in terms of public feedback.

Figure 34. Proposed project community feedback scores

4.3 Equity Impact Quantification
A ModelBuilder tool was developed to assign each proposed project with a score that
quantified how well it addressed geographic areas that have been historically underserved,
socially and economically, using input from large community stakeholders (C/CAG, SamTrans,
MTC EPC, and CAHPI as described in section 3.4). Each proposed project received a score of
104
zero, one, two, three, or four. One point was given for each equity polygon dataset that the
proposed project intersected. Figure 35 shows the equity score of each proposed project, derived
from the number of community stakeholder equity polygon datasets each proposed project
intersected with.

Figure 35. Proposed project equity scores

Proposed projects that score higher in equity impact have more avenues to external
funding and help push the equity-related agendas of community stakeholders. Proposed projects
that scored a four are concentrated on the eastern side of the city. The proposed projects on the
western side of the city scored lower with many projects receiving a score of zero.
105
4.4 Site Selection
Following the development of a bikeability score that was interpolated from street
segments to each of the proposed projects, a score that quantified workshop feedback, and a
score that describes the extent to which each proposed project addresses issues of equity, the site
selection process was undertaken. A weighted sum methodology was used to compute a final
score that would rank the proposed projects in order of implementation priority. Figure 36 shows
the location of each proposed project as well as how it ranked in terms of prioritization score.

Figure 36. Proposed project prioritization scores

Table 13 shows the bikeability, equity, workshop, and priority scores of the top-15
ranking proposed projects in terms of priority score.
106
Table 13. Fifteen highest-scoring proposed projects in terms of prioritization
Project Name
Bikeability
Score
Equity
Score
Workshop
Score
Priority
Score
59 - Redwood Ave. 2.36 4 2.25 2.86
74 - Vera Ave. 2.09 4 2.50 2.86
113 - El Camino Real 1.98 4 2.60 2.86
82 - Hudson St. 2.47 4 2 2.82
72 - Poplar Ave. 2.40 4 2 2.79
122 – Maple St. 2.36 4 2 2.78
37 – Middlefield Rd. 2.06 4 2.25 2.76
120 – Middlefield Rd. 2.06 4 2.25 2.76
121 – Woodside Rd. 2.05 4 2.17 2.73
12 – Path through Red
Morton Park
2.80 3 2.25 2.68
111 – Middlefield Rd. 1.68 4 2.25 2.64
8 - US 101 overcrossing at
Haven Ave.
3.76 4 0 2.56
17 - Maple St. 1.50 4 2 2.50
24 – Marsh Rd. 3.17 4 0 2.35
78 – Second Ave. 2.86 4 0 2.27

The project goal was to create design proposals for the three highest-ranking projects in
terms of priority score. Redwood Avenue, Vera Avenue, and El Camino Real scored the highest.
However, it would be redundant to make designs for proposed projects that already have designs
completed by the city and cycling infrastructure improvements along El Camino Real have
already been investigated during the development of the El Camino Corridor Plan (City of
Redwood City 2017). As such, it was disregarded from this study and replaced by the fourth-
highest scoring project, Hudson Street. Ultimately, Redwood Avenue, Vera Avenue, and Hudson
107
Street were selected as the proposed projects to be addressed by the design phase of the project.
The locations of these three proposed projects are shown in Figure 37.

Figure 37. The three proposed projects selected to receive design proposals

4.5 Proposed Project Design Proposals
In this section, the final designs for each of the three projects are described and
justifications for the design decisions are provided. For each of the three projects, there are three
maps that illustrate the surrounding context, a StreetMix design for the street as it appears in the
present, a StreetMix design for the new street design, and two renderings of the final SketchUp
model.
108
4.5.1 Redwood Avenue
The Redwood Avenue proposed project addresses a segment of Redwood Ave. that runs
3,971 feet from Virginia Avenue at the southwest to Ebener Street at the northeast and is
approximately 60 feet wide, including sidewalks. Redwood Ave. runs parallel to Oak Avenue,
which is one block to the northwest. Roosevelt Avenue also runs parallel to Redwood Ave. and
is located two blocks to the northwest. Roosevelt Ave. is considered an arterial route and
receives considerably more automotive use than Redwood Ave. The site is shown in Figure 38.
Redwood Avenue proposed project site

Figure 38. Redwood Avenue proposed project site

109
The northeast terminus of the Redwood Ave. proposed project is near a number of
amenities including Hawes Elementary School, Hawes Park, a church, Palm Park, the Sequoia
YMCA, and dining and retail along Woodside Road. The southwest terminus of Redwood Ave is
two blocks from Roosevelt Center, which is a shopping center featuring dining, retail, groceries,
and more. The Roosevelt Center is also adjacent to a church, public library, and Roosevelt
Elementary School. Improved cycling infrastructure along the entire extent of the Redwood Ave.
proposed project would make these amenities, which can be seen in Figure 39, easier to access
using non-car means of transportation.

Figure 39. Amenities near the Redwood Avenue proposed project

110
Nearly all of the parcels along the proposed Redwood Ave. project are zoned for duplex
residential use (Figure 40). While more densely populated than single-family zoning, duplex-
heavy neighborhoods are not as densely populated as multi-family zoned areas. Most duplex
homes do not have parking garages like multi-family buildings, and ample street parking may
still be required. Additionally, many of the homes along this portion of Redwood Ave. are
single-family homes, despite the duplex designation.

Figure 40. Zoning designations near the Redwood Avenue proposed project

Figure 41 shows a StreetMix visual of Redwood Ave. as it appeared in 2023 (Figure 41,
top) and the proposed design with an emphasis on improving bikeability (Figure 41, bottom).
111
The redesign features a two-way bike lane that is protected from traffic by a 0.6m wide curb and
a car parking lane.

Figure 41. Redwood Ave. illustration of 2023 (top) and as proposed (bottom)

Figure 42. Redwood Avenue redesign aerial view shows a rendering of the Redwood
Avenue redesign made in SketchUp.
112

Figure 42. Redwood Avenue redesign aerial view

Figure 43. Redwood Avenue redesign perspective view shows an eye-level rendering of
the Redwood Avenue redesign made in SketchUp.
113

Figure 43. Redwood Avenue redesign perspective view

The Redwood Ave. redesign shifts the road from a two-way street to a one-way street.
There is precedent for this type of infrastructure change, exemplified by the western end of
Colorado Avenue in Santa Monica, CA. According to reports about the Colorado Ave. shift to
one-way traffic, “The [traffic] study acknowledged that one-way travel would ‘result in
potentially additional traffic redistribution.’ But it found ‘these traffic shifts can be fully
accommodated by the given traffic capacity [on] parallel corridors, without creating significant
operational issues or travel delays’” (Chandler 2013). As mentioned, Redwood Ave. runs parallel
to Oak Ave. and Woodside Road and Roosevelt Ave., which already receive more use. Therefore
sufficient options should remain for drivers who may have previously driven in both directions
along Redwood Ave.
The designed one-way street would run from the northeast towards the southwest, which
was an intentional decision. First, it funnels cars towards the Roosevelt Center, which is one of
114
the few commercial areas in Redwood City not located on or near El Camino Real or Woodside
Road. More use of the Roosevelt Center’s facilities could inspire future growth around it,
resulting in a walkable community. Additionally, restricting travel towards the Roosevelt Center
may result in economic losses for the businesses that use that space.
In the Redwood Ave. redesign, the existing sidewalks and sidewalk buffers would be
maintained. The bike lanes would be located on the southeast side of the street and would be
made up of a two-way Class IV bike lane, with each lane occupying nearly six feet of space. The
bike lane is painted green to signify its use, and the green coloring extends through intersecting
roads and driveways. A concrete curb would be constructed to provide protection for the bikers
and would be almost 2 feet wide and 6 inches high. The rest of the street would consist of a 7.2-
foot-wide parking lane on either side of an 11-foot-wide driving lane, which physically permits
access by emergency vehicles.
4.5.2 Vera Avenue
The Vera Avenue proposed project addresses the segment of Vera Ave. that runs 3,557
feet from Red Morton Park at the southwest to El Camino Real at the northeast and is
approximately 60 feet wide, including sidewalks. At the southwest terminus, there is a paved
bike and pedestrian path that provides access into Red Morton Park. Across El Camino Real at
the northeast terminus is Maple Street, which features a Class IV bike lane. The site is shown in
Figure 44. Vera Avenue proposed project site.
115

Figure 44. Vera Avenue proposed project site

The immediate area around the northeast end of Vera Ave. features an abundance of
mixed-use facilities, both existing and under construction as of 2023. Additionally, one of
Redwood City’s three existing Class IV bike lanes, as of 2023, is on Maple Street, quite close to
Vera Ave. Just beyond the immediate vicinity of the mixed-use facilities is the Redwood City
CalTrain and SamTrans station as well as downtown Redwood City. Furthermore, Vera Ave. is
mere blocks away from schools such as John Gill Elementary School, Hawes Elementary
School, and Sequoia High School. A redesign of the street’s infrastructure that maximizes
bikeability would make it theoretically possible for residents of the neighborhood to have safe
116
and efficient access to schools, park space, retail, and public transportation and accomplish daily
business without using a car. Figure 45 visualizes the existing amenities in the area surrounding
the proposed Vera Ave. project.

Figure 45. Amenities near the Vera Avenue proposed project

City zoning designations have a significant impact on the design of the Vera Ave. cycling
infrastructure. Zoning for the Vera Ave. study area is shown in Figure 46. The northern half of
the Vera Ave. proposed project is flanked by medium density multi-family residential zoning.
The southern half is flanked by duplex residential housing. These zoning designations imply a
relatively high density. Multi-family housing also often implies the existence of sufficient on-site
parking, whether as a garage or large driveway. While as of 2023, many of the parcels zoned for
117
multi-family housing feature single-family homes, an increase in population and development
will likely result in more multi-family options on Vera Ave. For this reason, the decision to
remove street parking from one side of the street was made. This decision is supported by the
fact that improved cycling infrastructure, as well as the fact that there is potential for a highly
bikeable route between Vera Ave. and the Redwood City CalTrain station, should promote
cycling as a primary means of transportation over driving and reduce the need to own a car.

Figure 46. Zoning designations near the Vera Avenue proposed project

Figure 47 shows a StreetMix illustration of Vera Avenue as it appeared in 2023 (Figure
47, top) and a StreetMix illustration of the redesign of Vera Ave. with an emphasis on improving
118
bikeability (Figure 47. Vera Avenue redesign in StreetMix, bottom). The redesign includes a
two-way protected bike lane with protection coming from a 0.4m wide curb and a parking lane.

Figure 47. Vera Avenue redesign in StreetMix

119
Figure 48. Vera Avenue redesign aerial view shows a rendered view of the Vera Avenue
redesign from above.

Figure 48. Vera Avenue redesign aerial view

Figure 49. Vera Avenue redesign perspective view shows an eye-level rendering of the
Vera Avenue redesign created in SketchUp.
120

Figure 49. Vera Avenue redesign perspective view

In the Vera Ave. redesign, the sidewalks and sidewalk buffers on both side of the street
are maintained. The street parking on the southeast side of the road is replaced by a two-way,
fully-protected, green-painted bike lane, with each lane traveling in opposite directions and
taking up just over 5.5 feet of width. Adjacent to the bike lane is a concrete curb with a width of
1.3 meters and a height of six inches. This barrier, albeit not tall, provides a clear divider
between bike space and car space. Next to the barrier is space for street parking. The designated
street parking space takes up 7.2 feet of width and is accessible to cars traveling towards El
Camino Real. The street parking was chosen to remain on the southeast side of the street as
parked cars can complement the concrete curb as a means of providing cyclists with physical
protection. Furthermore, parked cars will be facing northeast while cyclists in the adjacent lane
will be traveling southwest. This creates an opportunity for direct eye contact between cyclists
and parked drivers, increasing awareness in both the driver and cyclist and reducing the risk of a
121
parked car’s door opening suddenly in front of the cyclist. The rest of the street is made up of
two automotive lanes, one traveling in each direction. Each lane is just under 10 feet wide, and
there is sufficient space for emergency vehicles to continue to use Vera Ave. While this means
that cars will be driving right next to the sidewalk, pedestrians should be able to remain safe
given the sidewalk buffer, inclusion of traffic calming measures such as speed bumps, and the
option to walk on the southeast sidewalk. This design is intended to run the entire extent of Vera
Ave., providing access into Red Morton Park as well as to El Camino Real. In the future, cycling
infrastructure could be implemented across El Camino Real and extend connectivity to the
preexisting Class IV bike lanes on Maple Street.
4.5.3 Hudson Street
The Hudson Street proposed project addresses a segment of Hudson St. that runs 562 feet
from Poplar Avenue to the northwest to Palm Avenue to the southeast and is approximately 60
feet wide, including sidewalks. While this project is relatively short, it serves as a continuation of
infrastructure upgrades on segments of Hudson St. to the west. The site is shown in Figure 50.
122

Figure 50. Hudson Street proposed project site

The proposed project is one block to the west of Woodside Road, which is a popular
artery for automotive travel and features significant retail and dining opportunities. The stretch of
Hudson St. addressed by this proposed project contains Palm Park, which features a large grass
area and a play structure, as well as the Sequoia YMCA. Both of these amenities are on the
northern side of the street. One block to the west is a church, while two blocks to the west are
Hawes Park and Hawes Elementary School. Figure 51 shows some of the amenities near the
Hudson St. proposed project.
123

Figure 51. Amenities near the Hudson Street proposed project

In terms of zoning, the Hudson Street proposed project is flanked by a mixture of
residential duplex and low-density multi-family residential designations, as seen in Figure 52.
There are three lots zoned for low density multi-family residential along the southern edge of this
segment of Hudson St. and each is built up quite extensively while also featuring on-site parking.
However, despite the dense population surrounding this proposed project, all street parking was
maintained.
124

Figure 52. Zoning designations near the Hudson Street proposed project

Figure 53 shows a StreetMix illustration of Hudson Street as it appeared in 2023 (Figure
53, top) and a StreetMix illustration of the Hudson St. with an emphasis on improving bikeability
(Figure 53, bottom). The redesign includes a two-way bike lane with physical protection coming
from parked cars. There is also room for bollards to be implemented, which is shown in the
renderings.
125

Figure 53. Hudson Street redesign in StreetMix

Figure 54. Hudson Street redesign aerial view shows an aerial rendering of the Hudson
Street redesign created in SketchUp.
126

Figure 54. Hudson Street redesign aerial view
Figure 55. Hudson Street redesign perspective view shows an eye-level perspective of the
Hudson Street redesign created in SketchUp.

Figure 55. Hudson Street redesign perspective view
127

In the Hudson St. redesign, only the northern sidewalk and sidewalk buffer are preserved.
The southern sidewalk is preserved; however, its buffer was removed to make more space for the
bike infrastructure. Because the amenities found on this block, Palm Park and the Sequoia
YMCA, are on the northern side of the street, the decision was made to place both directions of
bike lanes on the northern side of the street. The two-way Class IV bike infrastructure is made up
of one lane going in each direction, with each lane occupying nearly six feet of width.
Immediately to the south of the bike infrastructure is a seven-foot parking lane. It would
likely be possible to install bollards to provide physical protection for cyclists, but parked cars
are equally, if not more, protective. Similarly to the Vera Ave. design, parked cars and cyclists
will be facing opposite directions, minimizing the risk of cyclists being hit by suddenly opening
car doors. To the south of the first parking lane are two lanes for automotive travel, one going in
each direction. Each of these lanes is 10 feet wide. Finally, to the south of these lanes is another
parking lane, which is also seven feet wide.
128
Chapter 5 Conclusions
A geodesign methodology was applied to support the advancement of bicycle infrastructure in
Redwood City, California. As geodesign is a complex multidisciplinary and iterative practice,
several research questions were answered and ultimately woven together to produce a set of final
output. This chapter summarizes the results and key takeaways of each component of the project,
while also emphasizing the utility of an interdisciplinary geodesign methodology for evaluating
and advancing bikeability. In addition, considerations will be made to the limitations of this
project as well as what future work, both in Redwood City and more broadly, may entail.
5.1 Bikeability Quantification
Quantifying bikeability for each street segment in Redwood City yielded interesting
results. In general, bikeability was found to be between low and moderate. However, there are a
few neighborhoods which support moderate to high levels of bikeability. Furthermore, there is
disconnect between Redwood Shores and the rest of Redwood City, which may dissuade
commuters from travelling between the two areas on bike. While Redwood Shores features high
levels of bikeability, it is isolated and realistically only provides significant utility and value to
intra-Redwood Shores cyclists. Redwood City’s bike network has significant opportunities to
expand, which can be done by connecting the communities that already support high bikeability
with improved infrastructure.
The bikeability quantification is validated by comparing the results of the Redwood City
analysis with the results of Winters’ 2013 study, which is outlined in section 2.3.2. Unlike this
project, Winters used a weighted overlay technique which resulted in a continuous raster that
quantified bikeability at any given location within the Vancouver study area. However, Winters
129
used similar input variables, including topography, destination density, and bike route access.
Winters’ results matched up with the results of this paper, indicating that bikeability is often
higher in downtown areas where there is already existing cycling infrastructure as well as an
abundance of locations worth cycling to, such as stores and restaurants. In addition, low
bikeability can often be found in areas with significant hilly terrain and minimal destination
opportunities. While the methodology used by Winters and in this study differed slightly, the
results do in fact validate each other.
Despite the success of the bikeability evaluation, it is important to consider a possible
limitation. It is challenging to ever be able to confirm whether the ideal configuration of weights
and scaled values was used in a GIS multi-criteria decision analysis, however referencing
literature, and soliciting feedback on the configuration and results can help optimize the weight
and scale decisions (Ryan and Nimick 2019).
5.2 Workshop
Conducting stakeholder engagement is a critical component of any geodesign project. It
is important to query the people of the place and find out what their wants and needs are for their
area. Failing to do so may result in poor designs or underutilized public spaces.
The workshop for this project was well-planned and ran smoothly, resulting in valuable
input from the local stakeholders who were in attendance. Throughout the workshop, the nine
participants who attended the entirety of the session remained engaged and demonstrated
thoughtful and collaborative thinking. 24 proposed projects were mentioned as feasible locations
for improved bikeability and while 24 proposed projects may appear low given the number of
total streets in Redwood City, one explanation is that the participants were seemingly united in
terms of identifying roads that they believed would be useful as cycling corridors. This is further
130
suggested by the fact that many of the streets had multiple mentions. Moreover, the relatively
small number of participants as well as the length of the workshop (1.5 hours) may have
contributed to the small number of mentioned streets.
In addition, it seemed as if all the participants were extremely pro-cycling and therefore it
was impossible to facilitate discussions with significant disagreement. Nobody voiced opinions
in favor of cars or against a complete overhaul in favor of cycling infrastructure, and it is likely
that some residents of Redwood City do share that sentiment. Workshop bias could have been
prevented by targeting more people from broader backgrounds. But overall, the workshop proved
successful and was an integral part of this thesis.
5.3 Equity Analysis
A GIS methodology was used to evaluate the extent to which each proposed project
would interact with communities that have historically been underserved, as evident by their
demographics. This was done by acquiring four sets of polygon spatial data, each created by a
different Bay Area stakeholder authority. Proposed projects were measured for the number of
datasets that they overlapped, with a higher number of datasets indicating a stronger impact on
resolving issues related to equity. Furthermore, a higher number of overlaps can also lead to an
increase in the amount of funding that a proposed project receives from external sources. The
methodology applied to conduct the equity analysis built off the work undertaken by Grisé and
El-Geneidy (2018). Like their work, this thesis ensured the inclusion of equity considerations in
the decision-making process regarding the placement of cycling infrastructure by using
polygonal spatial data that represented disadvantage communities, based on various
sociodemographic metrics.
131
This process was straightforward and yielded results that were to be expected based on
previous investigations into the dispersion of demographic information throughout the city. Most
of the equity polygons were located on the eastern side of the city, where median household
income is relatively low and non-white percentage is relatively high, when compared to the
western side of the city.
The relationship between equity and bikeability in Redwood City was further explored by
analyzing demographics in conjunction with the results of the bikeability analysis. This allowed
for an investigation into the relationship between bikeability and non-white percentage, zero-car
households, and median household income.
5.3.1 Bikeability and Demographics in Redwood City
A better understanding of bikeability in Redwood City was achieved by using a bivariate
symbology to view the bikeability results with demographic information at the block group scale.
Bikeability appears to be overall higher in block groups where the percentage of households that
do not own a car is higher, such as in neighborhoods 11, 12, and 14 (see Figure 56). This is a
positive result, as when the number of households that do not own a car is relatively high,
infrastructure that supports other means of transportation, such as cycling, should exist.
However, neighborhoods seven. eight, 15, and 16 feature a high zero car household percentage
but low bikeability (see Figure 56). These four neighborhoods are prime candidates to receive
infrastructure investments that improve bikeability as a high zero car household percentage
implies that the availability of other modes of transportation needs to be increased. The three
proposed projects that underwent the design stage of this thesis are all within or adjacent to
neighborhoods seven and eight (see Figure 56).
132

Figure 56. Bikeability and zero car household percentage in Redwood City

When comparing bikeability with median household income in Redwood City, a small
correlation becomes apparent. While there is not a major relationship between overwhelmingly
high bikeability and a high median household income, some of the higher income block groups
on the western side of the city have slightly better bikeability. This is especially evident in
neighborhoods three, nine, and 10 (see Figure 57). Alternatively, aside from neighborhood 14,
most of the block groups that have the lowest median household income feature bikeability
ranging from moderate to low (see Figure 57). The three proposed projects that received designs
133
in this project are all within neighborhoods that feature a low median household income and low
to moderate bikeability (see Figure 57).

Figure 57. Bikeability and median household income in Redwood City

The relationship between non-white percentage and bikeability (Figure 58) is similar to
the relationship between median household income and bikeability. Neighborhoods one, two,
three, five, and nine have a lower non-white percentage and have moderately higher bikeability
than other parts of the city. Neighborhoods six, eight, 15, and 16 have a higher non-white
percentage and have generally moderate to poor bikeability, Neighborhood 14, where a Class IV
bike lane was installed in 2022, has a high non-white percentage and a high level of bikeability.
134

Figure 58. Bikeability and non-white percentage in Redwood City

5.3.2 Bikeability and Demographics in Redwood Shores
In contrast to Redwood City, block group demographics in Redwood Shores demonstrate,
to some extent, homogeneity. There is minimal correlation between the small variation in
demographics and bikeability, which makes sense given that the community was master planned,
and its demographics had little, if any, impact on the allocation and distribution of infrastructure.
There is not much variation in terms of households that do not own a car and the percent of zero
car households in Redwood Shores ranges is fairly low (Figure 59).
135

Figure 59. Bikeability and zero car household percentage in Redwood Shores

Redwood Shores residents generally have a moderate to high median household income,
aside from the furthest south block group (Figure 60).
136

Figure 60. Bikeability and median household income in Redwood Shores

Redwood Shores generally has a moderate to high non-white percentage, with the highest
non-white percentages being located in the block groups that line the southeastern edge of the
community (Figure 61).
137

Figure 61. Bikeability and non-white percentage in Redwood Shores

The high bikeability of Redwood Shores makes sense given the appearance of the two
highest-weighted variables within the community: bike lane access and slope. Redwood Shores
features significant Class I bike infrastructure which greatly improves the overall bikeability of
the community’s streets. Redwood Shores is also flat, meaning that hilly terrain is a non-issue for
local cyclists.
5.4 Site Selection
Each proposed project was assigned a score for bikeability, public opinion, and equity
impact. These three scores were then merged in a weighted sum, resulting in a final total score
138
for each proposed project. This total score was used to rank projects in terms of implementation
priority. Projects that received a higher priority score were located on a road that as of 2022 had
poor bikeability, had strong resident support to be made more bikeable, and was located within
neighborhoods that have historically faced inequality in terms of social, economic, and transport
resources.
The results of the site selection calculations mirrored what would be expected based on
the locations of proposed projects that have a high impact on equity. There is a clear divide,
where proposed projects that ranked in the upper 50th percentile of priority are concentrated on
the eastern side of the city while proposed projects that ranked in the lower 50th percentile of
priority are concentrated on the western side of the city. Furthermore, proposed projects that
were mentioned during the workshop that also have a strong impact on equity make up most of
the projects between the 75th and 100th percentile. The three highest-scoring projects all either
intersected or were directly adjacent to all four of the proposed projects. The relationship
between the three proposed projects and the C/CAG Equity Focus Areas (Figure 62, top left),
SamTrans Equity Planning Areas (Figure 62, top right), MTC EPCs (Figure 62, bottom left), and
CAHPI polygons (Figure 62, bottom right) are visualized in Figure 62.
139

Figure 62. Relationship between the designed projects and equity polygons

The results of the site selection process validate the methodology, because proposed
projects which clearly address poor bikeability, inequality, and the needs of residents all scored
amongst the highest.
5.5 Urban Design Modeling
Each of the three selected sites underwent a methodical design process, resulting in 3D
models and renderings that visualized what improved cycling infrastructure could look like.
While the designs may be considered aggressive given the amount of construction they would
140
require, there is precedent for the changes, such as Colorado Avenue in Santa Monica.
Additionally, they would most likely significantly increase bicycle ridership given the safety
guaranteed by the protected bike lanes. Fully protected bike lanes were selected for all three sites
in accordance with the findings of DiGioia (2017), Thomas and DeRobertis (2013), and Standen
et al. (2021). All three designs work to improve bikeability, reduce transportation inaccessibility
in communities that have been historically underserved, and align with the voiced desires of
community members. The final designs accurately portray what the proposed infrastructure
could look like in terms of street space allocation, material colors, and a bolstered urban tree
canopy.
In addition, all three designs are practical and would increase cycling in the areas
surrounding them. Using the fully protected Redwood Avenue cycling infrastructure, parents
could escort or send their children to class at Hawes Elementary School on bike before heading
to the Roosevelt Center, again on bike, to pick up groceries. They could then cycle to pick up
their children and take them to Red Morton Park. This scenario would only require the parents to
cycle along seven blocks of unprotected cycling infrastructure throughout their entire day, which
minimizes risk and increases cycling rates. Residents living along the new Vera Avenue cycling
infrastructure could in theory engage in physical activity or recreation at Red Morton Park, cycle
to class, buy groceries, purchase clothing, eat out at a restaurant, and have straightforward access
to public transportation without owning a car. Citizens living along Hudson Street would be able
to use the new cycling infrastructure to spend an afternoon relaxing at Palm Park, use the gym at
the YMCA, or shop along Woodside Road without much worry of being hit by a car, as a result
of the fully protected cycling infrastructure.
141
5.6 Overall Utility of Methods
This project serves as a case study for how a geodesign methodology can be applied to a
study area to advance bikeability within that study area. It demonstrates the value in a mixed
methods approach and shows that combining techniques from multiple disciplines can have
impressive results. Incorporating multiple formats of data can lead to a better-informed decision-
making process.
This project ties in methodologies applied in previous bicycle-infrastructure-related
studies to produce an interdisciplinary roadmap that can be used to evaluate the existing state of
bicycle infrastructure and strategically and objectively plan out how to make improvements to
bicycle infrastructure. Existing literature was used to guide the bikeability quantification
(Arellana et al. 2020; Codina et al. 2022; Grigore et al. 2019; Grisé and El-Geneidy 2018; Krenn,
Oja, and Titze 2015; Olgun 2020; Porter et al. 2020; Winters et al. 2013) and the workshop
planning and application (Arellana et al. 2020; Mueller et al. 2018; Mahyar et al. 2016). In the
future, this project can serve as an aggregation of past work while also offering new insights as
to how to go about improving the bikeability of a given study area.
However, there are important details to consider before attempting to apply this
methodology to a study area. First, relevant data must exist and be publicly available. While the
data does not have to be identical to the six variables used in this study, it should still be relevant
to bikeability and supported by the body of literature. Assuming spatial data is available, a GIS is
required. While ArcGIS Pro was used for this project, free GIS software such as QGIS also
exists and could be used. SketchUp Pro was used in this project to create the design mockups,
however there is a free version of SketchUp that would still prove useful in creating street
142
designs. In addition, SketchUp offers a one-month free trial for the Pro version, which is an
opportunity that can be taken advantage of to conduct a similar design process.
While these software and their tools used in this thesis are not overly-complicated, it is
still likely that a spatial scientist or GIS analyst would be required to undertake the project.
Fortunately, many cities now employ GIS analysts who would be capable of applying this
methodology to their jurisdiction.
Specific aspects of the methodology can also be further investigated to optimize the
workflow. For example, an analysis comparing the use of different weights in the bikeability
weighted sum analysis could be performed to determine the optimal configuration to assess
bikeability. The accuracy and effectiveness of this methodology would undoubtedly improve
because of learning experiences if it were applied repeatedly in different geographic
environments.
Overall, geodesign methodologies are tremendously valuable tools that spark
collaboration and innovation and show immense promise as a discipline that can bring about
high-quality solutions to pressing humanitarian issues in the built and natural environment, as
made evident by this thesis investigation of bikeability in Redwood City.
143
References
Arellana, J., Saltarín, M., Larrañaga, A. M., González, V. I., and Henao, C. A. 2020.
“Developing an Urban Bikeability Index for Different Types of Cyclists as a Tool to
Prioritise Bicycle Infrastructure Investments.” Transportation Research Part A: Policy
and Practice 139 (September): 310–34. https://doi.org/10.1016/j.tra.2020.07.010.
Braun, L. M., Rodriguez, D., A., and Gordon-Larsen, P. “Social (in)Equity in Access to Cycling
Infrastructure: Cross-Sectional Associations between Bike Lanes and Area-Level
Sociodemographic Characteristics in 22 Large U.S. Cities.” Journal of Transport
Geography 80 (October 1, 2019): 102544.
https://doi.org/10.1016/j.jtrangeo.2019.102544.
California Department of Housing and Community Development n.d. “Prohousing Designation
Program.” CA.Gov. Accessed 2023. https://www.hcd.ca.gov/planning-and-community-
development/prohousing-designation-program.
California Open Data Portal 2019. “CA Geographic Boundaries.” CA.gov. Last modified
October 23, 2019. Accessed 2022. https://data.ca.gov/dataset/ca-geographic-boundaries.
Campagna, M. “Metaplanning: About Designing the Geodesign Process.” Landscape and Urban
Planning, Geodesign—Changing the world, changing design, 156 (December 1, 2016):
118–28. https://doi.org/10.1016/j.landurbplan.2015.08.019.
C/CAG of San Mateo County n.d. “Countywide Bicycle and Pedestrian Plan.” City/County
Association of Governments of San Mateo County. Accessed 2023.
https://tooledesign.github.io/F0066-San-Mateo-CCAG/.
Chandler, J. “City: One-Way Car Flow on Colorado Is Better for Traffic.” Patch. March 15,
2013. Accessed 2023. https://patch.com/california/santamonica/city-one-way-car-flow-
on-colorado-is-better-for-traffic.
Chang, Y, Schau, E.M., and Finkbeiner, M. n.d. “Application of Life Cycle Sustainability
Assessment to the Bamboo and Aluminum Bicycles in Surveying Social Risks of
Developing Countries,” 24.
City of Redwood City 2017, “El Camino Corridor Plan”, City of Redwood City, CA. Accessed
2023. https://www.redwoodcity.org/departments/community-development-
department/planning-housing/planning-services/general-plan-precise-plans/el-camino-
real-corridor-plan.
City of Redwood City 2022, “RWC Walk Bike Thrive”, City of Redwood City, Redwood City,
CA. Accessed 2023. https://www.redwoodcity.org/departments/community-
development-and-transportation/engineering-transportation/transportation-and-
parking/rwc-walk-bike-thrive.
144
Codina, O., Maciejewska, M., Nadal, J., and Marquet, O. “Built Environment Bikeability as a
Predictor of Cycling Frequency: Lessons from Barcelona.” Transportation Research
Interdisciplinary Perspectives 16 (December 1, 2022): 100725.
https://doi.org/10.1016/j.trip.2022.100725.
Cottrill, C. D., and Thakuriah, P. V. "Evaluating pedestrian crashes in areas with high low-
income or minority populations." Accident Analysis & Prevention 42, no. 6 (2010):
1718-1728.
Cunha, I., and Silva, C. “Equity Impacts of Cycling: Examining the Spatial-Social Distribution
of Bicycle-Related Benefits.” International Journal of Sustainable Transportation 0, no. 0
(June 7, 2022): 1–19. https://doi.org/10.1080/15568318.2022.2082343.
Dave, S. n.d. “Life Cycle Assessment of Transportation Options for Commuters,” 16.
de, Hartog J. J., Boogaard, J., Nijland, H., and Hoek, G. 2010. “Do the Health Benefits of
Cycling Outweigh the Risks?” Environmental Health Perspectives 118 (8): 1109–16.
https://doi.org/10.1289/ehp.0901747.
DiGioia, J., Watkins, K. E., Xu, Y., Rodgers, M., and Guensler, R. “Safety Impacts of Bicycle
Infrastructure: A Critical Review.” Journal of Safety Research 61 (2017): 105–19.
Dill, J., and Rose, G. “Electric Bikes and Transportation Policy: Insights from Early Adopters.”
Journal of the Transportation Research Board 2314, no. 1 (2012).
https://journals.sagepub.com/doi/abs/10.3141/2314-01?journalCode=trra.
Duan, L. “Public Participation.” Lecture, University of Southern California, Los Angeles, CA,
September 1, 2021.
Fosgerau, M., Łukawska, M., Paulsen, M., and Rasmussen, T. J. “Bikeability and the Induced
Demand for Cycling.” Proceedings of the National Academy of Sciences of the United
States of America 120, no. 16 (April 18, 2023): e2220515120.
https://doi.org/10.1073/pnas.2220515120.
Galpern, P., Ladle, A., Uribe, F. A., Sandalack, B., and Doyle-Baker, P. 2018. “Assessing Urban
Connectivity Using Volunteered Mobile Phone GPS Locations.” Applied Geography 93
(April): 37–46. https://doi.org/10.1016/j.apgeog.2018.02.009.
Götschi, T., Garrard, J., and Giles-Corti, B. “Cycling as a Part of Daily Life: A Review of Health
Perspectives.” Transport Reviews 36, no. 1 (January 2, 2016): 45–71.
https://doi.org/10.1080/01441647.2015.1057877.
Grigore, E., Garrick, N., Fuhrer, R., and Axhausen, I., K., W. “Bikeability in Basel.”
Transportation Research Record: Journal of the Transportation Research Board 2673, no.
6 (June 2019): 607–17. https://doi.org/10.1177/0361198119839982.
145
Grisé, E., and El-Geneidy, A. “If We Build It, Who Will Benefit? A Multi-Criteria Approach for
the Prioritization of New Bicycle Lanes in Quebec City, Canada.” Journal of Transport
and Land Use 11, no. 1 (2018): 217–35.
Hinterthuer, A. “Study suggests trees are crucial to the future of our cities.” University of
Wisconsin-Madison. March 25, 2019. https://news.wisc.edu/study-suggests-trees-are-
crucial-to-the-future-of-our-cities/.
Hull, A., and O’Holleran, C. “Bicycle Infrastructure: Can Good Design Encourage Cycling?”
Urban, Planning and Transport Research 2, no. 1 (January 1, 2014): 369–406.
https://doi.org/10.1080/21650020.2014.955210.
Information Services. n.d. “GIS Data Download.” County of San Mateo. Accessed 2022.
https://www.smcgov.org/isd/gis-data-download.
Iracheta, M. “Redwood City among 11 New Prohousing Communities Designated by Gov.
Newsom.” Redwood City Pulse, April 6, 2023. https://www.rwcpulse.com/local-
news/housing/redwood-city-among-11-new-prohousing-communities-designated-by-gov-
newsom-6820899.
Jafari, Zahra, Bryan E. Kolb, and Majid H. Mohajerani. “Chronic Traffic Noise Stress
Accelerates Brain Impairment and Cognitive Decline in Mice.” Experimental Neurology
308 (October 1, 2018): 1–12. https://doi.org/10.1016/j.expneurol.2018.06.011.
Jehn, N., Atiquzzaman, M., and LaMondia, J. “A GIS Performance Measure-Based
Methodology for Prioritizing Bike Share System Expansion,” January 8, 2018.
Krenn, P., Oja, P., and Titze, S. “Development of a Bikeability Index to Assess the Bicycle-
Friendliness of Urban Environments.” Open Journal of Civil Engineering 05 (January 1,
2015): 451–59. https://doi.org/10.4236/ojce.2015.54045.
Kweon, B, Shin, W., Folzenlogen, R., Kim, J., and Texas Transportation Institute. “Children and
Transportation: Identifying Environments That Foster Walking and Biking to School,”
September 1, 2006. https://rosap.ntl.bts.gov/view/dot/40169.
Li, W., and Milburn, L. A. 2016. “The Evolution of Geodesign as a Design and Planning Tool.”
Landscape and Urban Planning 156 (December): 5–8.
https://doi.org/10.1016/j.landurbplan.2016.09.009.
Lindsey, G., Tao, T., Wang, J., and Cao, J. "Pedestrian and Bicycle Crash Risk and Equity:
Implications for Street Improvement Projects." (2019).
Loukaitou-Sideris, A., Liggett, R., and Sung, H. G. "Death on the crosswalk: A study of
pedestrian-automobile collisions in Los Angeles." Journal of Planning Education and
Research 26, no. 3 (2007): 338-351.
146
Ma, L., Ye, R., and Wang, H.. “Exploring the Causal Effects of Bicycling for Transportation on
Mental Health.” Transportation Research Part D: Transport and Environment 93 (April 1,
2021): 102773. https://doi.org/10.1016/j.trd.2021.102773.
Mahyar, N., Burke, K. J., Xiang, J., Meng, S., Booth, K. S., Girling, C. L., and Kellett, R. W.
"Ud co-spaces: A table-centred multi-display environment for public engagement in
urban design charrettes." In Proceedings of the 2016 ACM international conference on
interactive surfaces and spaces, pp. 109-118. 2016.
McNeil, N. 2011. “Bikeability and the 20-Min Neighborhood: How Infrastructure and
Destinations Influence Bicycle Accessibility.” Transportation Research Record 2247 (1):
53–63. https://doi.org/10.3141/2247-07.
McQueen, M., MacArthur, J., and Cherry, C. “The E-Bike Potential: Estimating Regional e-Bike
Impacts on Greenhouse Gas Emissions.” Transportation Research Part D: Transport and
Environment 87 (October 1, 2020): 102482. https://doi.org/10.1016/j.trd.2020.102482.
Monsere, C., Dill, J., McNeil, N., Clifton, K., Foster, N. “Lessons from the Green Lanes:
Evaluating Protected Bike Lanes in the U.S.” Portland State University Library, June
2014. https://doi.org/10.15760/trec.115.
MRLC Data n.d. “NLCD 2016 USFS Tree Canopy Cover (CONUS).” Multi-Resolution Land
Characteristics Consortium. Accessed 2022. https://www.mrlc.gov/data/nlcd-2016-usfs-
tree-canopy-cover-conus.
MTC GIS 2020. “Equity Priority Communities - Plan Bay Area 2050.” MTC GIS. Accessed
2023. https://opendata.mtc.ca.gov/datasets/equity-priority-communities-plan-bay-area-
2050/explore?location=37.878600%2C-122.370850%2C9.04.
Mueller, J., Lu, H., Chirkin, A., Klein, B., and Schmitt, G. "Citizen Design Science: A strategy
for crowd-creative urban design." Cities 72 (2018): 181-188.
National Map n.d. “TNM Download (v2.0).” USGS. Last modified August 11, 2022. Accessed
2022. https://apps.nationalmap.gov/downloader/.
Nielsen, T. A. S., Skov-Petersen, H., and Carstensen, T. A. 2013. “Urban Planning Practices for
Bikeable Cities – the Case of Copenhagen.” Urban Research & Practice 6 (1): 110–15.
https://doi.org/10.1080/17535069.2013.765108.
Oja, P., Titze, S., Bauman, A., de Geus, B., Krenn, P., Reger-Nash, B., & Kohlberger, T. (2011).
Health benefits of cycling: a systematic review. Journal of Physical Activity and Health,
8(8), 1153-1163.
Olgun, R. 2020. “Sustainable Bicycle Path Planning for Medium-Sized Cities by Using GIS-
Based Multicriteria Decision-Making Analysis: A Case Study From Turkey,” 10.
147
Olsson, L. E., Gärling, T., Ettema, D., Friman, M., and Fujii, S. 2013. “Happiness and
Satisfaction with Work Commute.” Social Indicators Research 111 (1): 255–63.
https://doi.org/10.1007/s11205-012-0003-2.
Parker, K. M., Rice, J., Gustat, J., Ruley, J., Spriggs, A., and Johnson, C. “Effect of Bike Lane
Infrastructure Improvements on Ridership in One New Orleans Neighborhood.” Annals
of Behavioral Medicine: A Publication of the Society of Behavioral Medicine 45 Suppl 1
(February 2013): S101-107. https://doi.org/10.1007/s12160-012-9440-z.
Porter, A., Kohl, H., Perez, A., Reininger, B., Gabriel, K., P., and Salvo, D. “Bikeability:
Assessing the Objectively Measured Environment in Relation to Recreation and
Transportation Bicycling.” Environment and Behavior 52, no. 8 (2020): 861–94.
https://doi.org/10.1177/0013916518825289.
Public Health Alliance of Southern California n.d. “California Healthy Places Index.” Public
Health Alliance of Southern California. Accessed 2023.
https://map.healthyplacesindex.org/?redirect=false.
Pucher, J., and Buehler, R. 2008. “Making Cycling Irresistible: Lessons from The Netherlands,
Denmark and Germany.” Transport Reviews 28 (4): 495–528.
https://doi.org/10.1080/01441640701806612.
Qin, J., Lee, S., Yan, X., and Tan, Y. "Beyond solving the last mile problem: the substitution
effects of bike-sharing on a ride-sharing platform." Journal of Business Analytics 1, no. 1
(2018): 13-28.
Redwood City GIS. n.d. “Redwood City Community GIS.” City of Redwood City. Accessed
2022. http://webgis.redwoodcity.org/community/.
Reggiani, G., van Oijen, T., Hamedmoghadam, H., Daamen, W., Vu, H. L., and Hoogendoorn, S.
"Understanding bikeability: a methodology to assess urban networks." Transportation 49,
no. 3 (2022): 897-925.
Ryan, S., and Nimick, E.. “Multi-Criteria Decision Analysis and GIS.” ArcGIS StoryMaps,
August 11, 2019.
https://storymaps.arcgis.com/stories/b60b7399f6944bca86d1be6616c178cf.
Saelens, B. E., Sallis, J. F., and Frank, L. D. 2003. “Environmental Correlates of Walking and
Cycling: Findings from the Transportation, Urban Design, and Planning Literatures.”
Annals of Behavioral Medicine 25 (2): 80–91.
https://doi.org/10.1207/S15324796ABM2502_03.
San Mateo County GIS n.d. “San Mateo County GIS Open Data.” County of San Mateo.
Accessed 2022. https://data-smcmaps.opendata.arcgis.com/.
ShockleyD_Samtrans 2022. “SamTrans Equity Planning Areas.” SamTrans. Last modified
February 11, 2022. Accessed 2023.
https://www.arcgis.com/home/item.html?id=43327d2f4cf44e39b002e179e28ea640.
148
Standen, C., Crane, M., Greaves, S., Collins, A. T., and Rissel, C. “How Equitable Are the
Distributions of the Physical Activity and Accessibility Benefits of Bicycle
Infrastructure?” International Journal for Equity in Health 20, no. 1 (September 15,
2021): 208. https://doi.org/10.1186/s12939-021-01543-x.
Staunton, C., Hubsmith, D., and Kallins, W. “Promoting Safe Walking and Biking to School:
The Marin County Success Story.” American Journal of Public Health 93, no. 9 (2003).
https://ajph.aphapublications.org/doi/full/10.2105/AJPH.93.9.1431.
StreetMix n.d. “StreetMix.” StreetMix. Accessed 2023. https://streetmix.net/.
Thomas, B., and DeRobertis, M.. “The Safety of Urban Cycle Tracks: A Review of the
Literature.” Accident Analysis & Prevention 52 (2013): 219–27.
Transportation Injury Mapping System n.d. “SWITRS Query & Map.” University of California,
Berkeley. Last modified 2022. Accessed 2022. https://tims.berkeley.edu/.
United States Census Bureau 2022. “QuickFacts: Redwood City, California.” United States
Census Bureau. Last modified July 1, 2022.
census.gov/quickfacts/redwoodcitycitycalifornia.
vbeckley_RWC 2019. “Neighborhoods Layer.” ArcGIS Online. Last modified August 20, 2019.
Accessed 2022.
https://uscssi.maps.arcgis.com/home/item.html?id=f53c61286280425190ec57076d05515
c.
Wardlaw, M. J. 2000. “Three Lessons for a Better Cycling Future.” BMJ 321 (7276): 1582–85.
https://doi.org/10.1136/bmj.321.7276.1582.
Winters, M., Brauer, M., Setton, E. M., and Teschke, K. 2013. “Mapping Bikeability: A Spatial
Tool to Support Sustainable Travel.” Environment and Planning B: Planning and Design
40 (5): 865–83. https://doi.org/10.1068/b38185.
Zhang, Y, and Mi, Z. “Environmental Benefits of Bike Sharing: A Big Data-Based Analysis.”
Applied Energy 220 (June 15, 2018): 296–301.
https://doi.org/10.1016/j.apenergy.2018.03.101.

Abstract (if available)

Abstract In the United States, emission-releasing cars reign as the leading form of transportation among citizens. Given the increasing effects of global climate change, it is critical that society finds alternative solutions to travel. The use of geodesign, combining data-driven spatial analysis with thoughtful design and community input, shows promise as an approach to better design transportation infrastructure in the US. This thesis applies a geodesign methodology to propose biking infrastructure improvements in Redwood City, CA. It first assesses existing bikeability as of 2022 using a spatial analysis in a GIS. It finds that Redwood City has moderate bikeability with potential for improvements that if implemented, will simultaneously help solve issues related to economic, social, and transport inequity. The project next selects three specific street segments in the city that could most benefit from improved biking infrastructure. It makes the selection through a combined analysis of these bikeability results, assessments by local stakeholder organizations of underserved areas, and community feedback gathered at a public workshop organized by the author that focused on biking in the city. These three selected street segments underwent a design process, resulting in models and renderings of what improved cycling infrastructure could look like in Redwood City. This thesis ultimately serves as an exemplar methodology that can be applied to other cities in the US to increase local bikeability and improve long-term sustainability in terms of social equity and the environment.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Rails to trails web mapping application for the great Redwood trails: mapping northern California’s repurposed trails

PDF

Human suffering during wartime: a StoryMap of violations of international law during the Russo-Ukrainian War

PDF

Development of a historical urban land use web application for the city of Hong Kong

PDF

That sinking feeling: predicting land subsidence in California’s San Joaquin Valley with a spatial regression model

PDF

Mapping punk music and its relative subgenres

PDF

Enriching the Demographic Survey sampling for the Los Angeles County Annual Homeless Count with spatial statistics

PDF

A spatial analysis of veteran healthcare accessibility

PDF

Exploring the pernicious effects of redlining and discriminatory policies on an American city: a spatio-temporal case study of New York City

PDF

Spatiotemporal analysis of the SLOSH and ADCIRC storm surge models: a case study of hurricane Ida

PDF

Alaska Hike Search: designing a mapping application for Alaskan trails and user contributed hazards

PDF

Generating trail conditions using user contributed data through a web application

PDF

Site suitability analysis for implementing tidal energy technology in southern California

PDF

The community wildfire protection plan repository: using VGI to create a national collection of wildfire management information

PDF

Silicon Valley construction project web mapping application

PDF

Pharmacy shortage areas across the United States: a visual representation by a web mapping application

PDF

Collecting and managing VGI infrastructure assessments in support of stability operations

PDF

Philly Bike Report: a mobile app for mapping and sharing real-time reports of illegally blocked bike lanes in Philadelphia

PDF

A comparison of pixel-based, object-based, deep learning, and data fusion image classification approaches in delineating urban tree canopy in City Terrace

PDF

Assessing homeless accessibility to community resources in the city of Los Angeles

PDF

A spatial and temporal exploration of how satellite communication devices impact mountain search and rescue missions in California’s Sierra Nevada mountain range

Asset Metadata

Creator Huisman, Erik Siebren (author)

Core Title Advancing Redwood City's bicycle infrastructure through a geodesign workflow

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Degree Conferral Date 2023-08

Publication Date 05/19/2023

Defense Date 05/09/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bicycle infrastructure,bikeability,geodesign,GIS,OAI-PMH Harvest,spatial sciences,Urban planning

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sedano, Elisabeth (committee chair), Duan, Leilei (committee member), Huang, Guoping (committee member)

Creator Email ehuisman@usc.edu,erikshuisman@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113132906

Unique identifier UC113132906

Identifier etd-HuismanEri-11873.pdf (filename)

Legacy Identifier etd-HuismanEri-11873

Document Type Thesis

Format theses (aat)

Rights Huisman, Erik Siebren

Internet Media Type application/pdf

Type texts

Source 20230522-usctheses-batch-1047 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu