Monitoring the
Compliance of Mobility
Service Providers
with Municipal Policy
Agendas:
a focus on the first and
the last mile
Francisco Edson Macedo Filho | S1040872
August 2020
Radboud University | Rebel Group
Master’s Program in Spatial Planning
Urban and Regional Mobility
Monitoring the compliance of mobility service providers
with municipal policy agendas:
a focus on the first and the last mile
_
Master’s thesis Spatial Planning
Author:
Francisco Edson Macedo Filho
Student number: s1040872
e-mail: francisco.macedo@rebelgroup.com
Master program:
Master Spatial Planning
Urban and Regional Mobility
University:
Radboud University Nijmegen
Faculty of Management Sciences
Nijmegen, the Netherlands
Supervisors/ readers:
1st: Kevin Raaphorst (PhD)
2nd: Sander Lenferink (PhD)
Mentor: Regina Liptak (Rebel Group | Transit and Ticketing Division)
_
Abstract
Emergent modes of transport, multimodal journey planning platforms,
shared mobility applications, faster forms of payment, etc; all increase
the number of options available to potential users, promoting seamless
travel experiences. They also lead to a better understanding of travel
patterns, and produce overwhelming amounts of data. With new smart
mobility solutions, new challenges arise: limited infrastructure to
accommodate the numerous “smaller-than-car” vehicles, disorganized
occupation of public spaces, traffic safety concerns, non-inclusive
spatial and financial distribution of mobility systems. A key priority for
policy makers and planners is to help ensure that these services are
accessible to a wide range of users, and are compliant to existing urban
mobility policies. As mobility options proliferate, policy frameworks
to govern mobility data and regulate private operators are overdue. In
that sense, how can the public sector (e.g. municipalities, agencies,
decision-makers) systematically evaluate the extent to which (micro-)
mobility service providers are aligned with operational and strategic
goals of cities? This research aims to develop and apply a systematic
approach to assist the public sector to evaluate private micro-mobility
services in relation to the extent to which they effectively comply with
strategic and operational policies, with focus on equity as an universal
policy goal To achieve the proposed aim, a set of indicators capable of
benchmarking compliance are proposed and applied to a group of case
studies in the city of Los Angeles.
Although the proposed framework can be
applied with reasonable success, several challenges exist (e.g. data availability).
However biased and limited in many aspects mentioned in this research,
the proposed approach can be flexible enough to support continuous
developments of mobility policies over time through the inclusion and
maintenance of strategic goals and indicators, as the knowledge about
the explored mobility issue is accumulated.
1.Research Framework
8
1.1. Research Context and Problem
9
1.2. Research Relevance
12
1.2.1. Societal Relevance
12
1.2.2. Scientific Relevance
13
2.Theoretical Framework
14
2.1. Literature Review and Theoretical Framework
15
2.2. Openness and Standardization of Mobility Data
15
2.3. Diving Deeper into Mobility Data Standards
17
2.4. Monitoring Urban Mobility: Frameworks and Indicators
21
2.5. Framing Policy Goals and Compliance of Private Mobility Providers
26
3.Research Design
28
3.1. Research strategy
29
3.2. Application of the Framework in a Case Study
30
3.2.1. Universal Strategic Goals in Transportation and Equity
32
3.3. Research Methods, Data Collection and Data Analysis
38
3.3.1. Evaluation Framework
38
3.3.2. Data Collection
43
3.3.3. Data Manipulation and Analysis
46
3.3.4. Compliance Assessment
48
3.3.5. Qualitative Validation and Analysis
48
4.Results 49
4.1. Proposed Framework
50
4.2. Mobility providers’ performance
57
4.2.1. Comparative Analysis [Level 4]
69
4.2.2. Criteria Weighting [Level 3]
76
4.2.3. Compliance Assessment [Level 2]
77
5. Reflections, Conclusions and Recommendations
80
5.1. Reflections
81
5.2. Conclusions
83
5.3. Recommendations for Future Planning Practices and Academic Studies
85
6.References 88
7.Annexes 95
List of Figures
Figure 1: Articles about mobility data and its usefulness for planning. Source: Own production.
10
Figure 2: Average trip duration versus transit ridership during covid-19. Source: bay wheels.
13
Figure 3: Standard-based APIs for transport operators to/ from MaaS providers. Source: Rijkswaterstaat (2019).
16
Figure 4: Non-exhaustive storyline of commercial and non-commercial mobility APIs. Source: Own production.
17
Figure 5: GTFS data feeds published worldwide. Source: Transitfeeds.com.
18
Figure 6: GTFS dataset structure. Source: google (2020).
18
Figure 7: GBFS data feeds published worldwide. Source: projected sample from github.com/NABSA/gbfs.
19
Figure 8: Key dimensions in building an indicator framework. Source: adapted from damidavicus et al. (2019).
21
Figure 9: Key dimensions in building an indicator framework. Source: adapted from Gudmundsson et al. (2016).
23
Figure 10: Typical structure of AHP. Source: adapted from Saaty (1980).
25
Figure 11: Conceptual framework and research operationalization. Source: Own production.
27
Figure 12: Methodological steps. Source: Own production.
29
Figure 13: The market of micro-mobility and data availability. Source: Own production.
31
Figure 15: Distribution of micro-mobility services and data providers in US. Source: Own production.
32
Figure 16: Analytic hierarchy process structure. Source: Own production.
38
Figure 17: Transit systems in los angeles (all modes). Source: Own production.
45
Figure 18: Administrative limits of california. Source: Own production.
45
Figure 19: Sample question posed to transportation experts of los angeles. Source: Own production.
46
Figure 20: Proposed approach for the analysis of spatial data. Source: Own production.
47
Figure 21: Proposed framework. Source: Own production.
50
Figure 22: Income of communities living within catchment area. Source: Own production.
54
Figure 23: Income of communities living within catchment area. Source: Own production.
55
Figure 24: Access costs to the nearest micro-mobility hubs. Source: Own production.
55
Figure 25: Number of transit hubs within catchment area of providers. Source: Own production.
56
Figure 27: Static location of the studied cases. Source: Own production.
57
Figure 26: Access costs to the nearest micro-mobility hubs. Source: Own production.
57
Figure 29: Share of potential users by income level. Source: own
58
Figure 28: Spatial distribution of jump. Source: own production.
58
Figure 30: Distribution of potential users by income level. Source: Own production.
59
Figure 31: Proportion of potential users by distance intervals [jump]. Source: Own production.
60
Figure 32: Nearest transit hubs by income level [jump]. Source: Own production.
60
Figure 34: Share of potential users by income level [la metro]. Source: Own production.
61
Figure 33: Spatial distribution of la metro. Source: own production.
61
Figure 35: Distribution of potential users by income level [la metro]. Source: Own production.
62
Figure 36: Proportion of potential users by distance intervals [la metro]. Source: Own production.
63
Figure 37: Nearest transit hubs by income level [la metro]. Source: Own production.
63
Figure 39: Share of potential users by income level [breeze]. Source: Own production.
64
Figure 38: Spatial distribution of breeze. Source: own production.
64
List of Tables
Figure 41: Accessibility indicators [jump]. Source: Own production
66
Figure 42: Nearest transit hubs by income level [breeze]. Source: Own production.
66
Figure 44: Share of potential users by income level [hopr]. Source: Own production.
67
Figure 43: Spatial distribution of hopr. Source: own production.
67
Figure 45: Normalized indicators chart. Source: Own production.
75
Figure 46: Consolidated results of the weighting process. Source: Own production.
76
Figure 47: Aggregated results (weighted and not weighted). Source: Own production.
79
Figure 48: Possible intervention towards more equitable supply. Source: Own production.
81
Figure 49: Round catchment areas versus isochrones. Source: Own production.
82
Figure 50: Possible intervention towards more equitable supply. Source: Own production.
87
Figure 51: Representation of minimum thresholds for a certain indicator. Source: Own production.
87
Table 1: Framing questions and answers for different planning frameworks. Source:gudmundsson et al.(2016).
22
Table 2: Mobility policies crafted by different californian municipalities. Source: Own production.
34
Table 3: Challenges and opportunities for micro-mobility. Source: fhwa (2017).
37
Table 4: An example of a technical report on the indicator. Source: adapted from mooney et al (2019)
40
Table 5: Scale for pairwise comparisons. Source: adapted from saaty and vargas (1991).
40
Table 6: Example of comparison matrix with 5 criteria (normalization). Source:: Own production.
41
Table 7: Example of comparison matrix with 5 criteria. Source: Own production.
41
Table 8: Random consistency table. Source: adapted from saary (1980).
42
Table 9: List of propposed indicators. Source: Own production.
53
Table 11: General statistics about the studied cases. Source: Own production.
57
Table 10: Access costs to the nearest micro-mobility hubs. Source: Own production.
57
Table 12: Accessibility indicators [jump]. Source: Own production
59
Table 13: Financial inclusiveness indicators [jump]. Source: Own production
60
Table 14: Integration with transit indicators [jump]. Source: Own production.
60
Table 15: Digital literacy indicators [jump]. Source: Own production.
60
Table 16: Accessibility indicators [la metro]. Source: Own production
62
Table 17: Financial inclusiveness indicators [la metro]. Source: Own production
63
Table 18: Integration with transit indicators [la metro]. Source: Own production.
63
Table 19: Digital literacy indicators [la metro]. Source: Own production.
63
Table 20: Accessibility indicators [breeze]. Source: Own production.
65
Table 23: Digital literacy indicators [breeze]. Source: Own production.
66
Table 22: Integration with transit indicators [breeze]. Source: Own production.
66
Table 27: Digital literacy indicators [hopr]. Source: Own production.
68
Table 26: Integration with transit indicators [hopr]. Source: Own production.
68
Table 24: Accessibility indicators [hopr]. Source: Own production
68
Table 25: Financial inclusiveness indicators [hopr]. Source: Own production.
68
Table 28: Accessibility indicators (combined analysis). Source: Own production.
69
Table 29: Accessibility indicators (combined analysis). Source: Own production.
70
Table 30: Financial inclusiveness (combined analysis). Source: Own production.
71
Table 31: Integration with transit (combined analysis). Source: Own production.
71
Table 33: Integration with transit (combined analysis). Source: Own production.
72
Table 32: Digital literacy (combined analysis). Source: Own production.
72
Table 34: Normalized indicators between providers. Source: Own production.
73
Table 35: Normalized indicators between providers. Source: Own production.
74
Table 36: Normalized indicators between providers. Source: Own production.
77
Table 37: Normalized scores of each criteria before weighting process. Source: Own production.
77
Table 38: Aggregated results (not weighted). Source: Own production.
78
1.
research
framework
According to Goonetilleke et al. (2014), the quantity of urban dwellers
has been increasing at an average rate of 60 million people/year during
the past decades. The negative side of that growth is the development
of largely unplanned and sprawled settlements, along with externalities
related to traffic and the environment. The ‘smart’ character of cities
emerges in a context where intense urbanization leads to many complex
problems (Dizdaroglu et al., 2012), challenging city administrations
and entrepreneurs to progressively provide essential services to urban
population, such as urban mobility, security, health, social equity to
services and amenities (Konys, 2018). Multimodal journey planning
platforms, shared mobility applications, electronic ticketing marketplaces,
faster forms of payment, etc; all increase the number of options available
to users, also leading, in theory, to a better knowledge of travel patterns
(Harbers, 2016), and an overwhelming amount of transport data being
produced and made available.
Micro-Mobility solutions (e-scooters, shared bikes), for instance, are
being regarded as the future of transportation in cities around the world,
with potential to create more sustainable travel behaviours (McKenzie,
2019). Given that around 60 percent of all trips in the United States
are five miles or less and that 46 percent of automobile trips are three
miles or less, there is an undeniable potential for “smaller-than-car”
transportation to enable more flexible local mobility while creating
greener, more connected and less congested cities (NHTS, 2018). And as
consumers take advantage of this growing trend, the market opportunity
continues to expand. Worldwide, investors have already poured more
than $5.7B into micro mobility startups during the past 4 years (McKinsey
& Company, 2019).
Along with new mobility solutions, new challenges arise: limited
infrastructure to accommodate the numerous “smaller-than-car”
vehicles, disorganized occupation of public spaces due to careless
drop-offs, safety concerns regarding the crescent number of users in cities not
yet prepared, unequal distribution of mobility, regarding to disadvantaged
communities (e.g. unbanked, unbanked, low-income) (Shaheen and
Cohen, 2019). In this sense, a key priority for policy makers, planners
and consultants is to help ensure that these services are accessible to
a wide range of population groups, and are compliant to existing and
urban mobility policies. In addition, new partnerships between public
and private parties can emerge in order to help municipalities to achieve
public goals. In the U.S, for example, cities have required that mobility
In parallel, this ecosystem is being extended to new types of consumers
- data consumers, such as data analysts, governmental agencies and
new entrepreneurs (Storme et al., 2020) through standardized , accurate
and open data formats, such as GTFS, GTFS-rt, GBFS, GBFS+, etc; thus
raising the awareness of society in general about the relevance of accurate
mobility data - on one hand, as a valuable Source of new revenue models;
and on the other hand, as relevant information concerning to future
decision-making regarding to urban planning (Harbers, 2016). Despite
various initiatives on promoting new mobility services, little is known
about how these systems are performing (Debnath, 2014) and using
the public space. Even less if they comply with public goals regarding to
urban mobility.
operators outline their plans for addressing spatial equity, developed
specific geographic requirements for public space occupation and
demanded up-to-date data generated by users (City Los Angeles, 2020;
City of Louisville, 2020; Washington D.C., 2020; etc). Other notorious
examples are startups offering advisory services to the public sector on
how to use standardized mobility data from private providers to manage
real-time traffic, enforce regulation (e.g. parking zones, low emission
areas), provide accurate travel information, understand and predict travel
behavior, etc (Figure 1).
F
igure
1:
recently
published
articles
about
mobility
data
and
its
useFulness
For
mobility
planning
. s
ource
: o
wn
production
.
According to Debnath (2014), without proper measurable indicators, it is
difficult to evaluate how well a transport system is performing in terms
of public goals and policies. The use of performance indicators to assess
transportation systems is among the best-practices worldwide and
reported in several urban mobility plans, policy documents and visions.
International cities such as London (Transport 2025), Paris (Plan de
Déplacements urbains ile-de-France), Singapore (Land Transport Master
plan), Toronto (The Big Move), Los Angeles (Mobility Plan 2035), Sydney
(Sydney Long Term Transport Master Plan), New York City (Plan 2040),
and many others propose objective measures to monitor how the built
environment and travellers are interacting with one another (Boisjoly and
Geneidy, 2017).
Although quite relevant, the mentioned indicators can be very challenging
to be generated due to poor availability of data and resources to enable
data collection (Boisjoly and Geneidy, 2017), or its concentration in the
hands of the companies willing to bring new solutions to the market
and trying to protect their business models (Harbers, 2016). Making
progress towards policy goals requires standards and metrics that can
continuously be monitored and evaluated. However, so far, there is no
wide range of methods and discussions available in the literature to assist
cities evaluate the performance and policy compliance by private mobility
service providers, as well as the quality of standardized data provided by
these actors to serve spatial planning purposes.
As mobility options proliferate, policy frameworks to govern mobility
data and regulate private operators are overdue. In that sense, how
can the public sector (e.g. municipalities, agencies, decision-makers)
systematically evaluate the extent to which (micro-) mobility service
providers are aligned with operational and strategic goals of cities?
The posed question can be divided in the following sub-questions:
(a) What types of systematic approaches can be applied by the public
sector to evaluate and monitor the performance of (micro-) mobility
providers?
(b) How can such approaches help to interpret mobility policies and use
their major elements to monitor micro-mobility operations in relation to
existing policy goals?
(c) What set of criteria and metrics can be proposed and measured
considering the application of the mentioned approach to operational
micro-mobility companies?
(d) Considering the proposed approach and its study through a case study,
to what extent are micro-mobility companies aligned with transportation
equity as an universal policy goal?
This research aims to develop and apply a systematic approach to assist
the public sector to evaluate private micro-mobility services in relation to
Municipalities are becoming progressively aware of how relevant it
is to understand how the public space is being used by new mobility
players in order to plan for the future of mobility in more data-driven
manners. However, as mentioned, most of the concern is allocated in the
operational level (e.g. fleet sizes, parking allowances, hours of operation),
as traffic management. In this study, it is believed that the scope of this
understanding can be broadened and contain longer-term perspectives,
with preoccupations correlated to strategic goals previously established
in mobility plans and visions. As many of these documents address
societal and environmental issues, such as transport equity, economic
vitality, climate change, etc; systematic manners to assist public
administrators to use produced to efficiently monitor how these issues
are being addressed by relevant players, can improve their capacity to
plan transport systems on behalf of society. From a market perspective,
the same is to be achieved, since this research aims also to reach private
mobility providers, raising their awareness about the advantages of
aligning their business models with resources and agendas provided
by the public sector - not only necessary regulatory basis to operate in
cities, but the existing infrastructure and strategic goals.
As the Coronavirus still develops across the globe, the way we travel
in cities might completely change for the long term, and gradually
remodeling the mobility industry. Travel might become more dependent
on personal transport modes, as many transit agencies have seen their
ridership numbers fall to levels below 70%. As cities slowly get back to
the new normal, certain modal shifts might already be happenning, as
the following figure suggests - demand for biking has shown first signs of
a long term shift towards longer trips using this mode. In San Francisco
(CA), according to data extracted from the Bay Wheels system, trips got
approximately 8 min (2,6km) longer 3 months after the establishment of
measures to restrict movements in the city.
Bike-share systems around the world are gaining popularity as
commuters fled transit systems. On the other hand, driving is also
rebounding all over the world, and could eventually return stronger than
1.2.
research
relevance
1.2.1.
Societal Relevance
the extent to which they effectively comply with strategic and operational
policies, with focus on equity as an universal policy goal. The city of Los
Angeles was chosen as scenario for this research due to the relatively
wide availability of micro-mobility open data, and the attention that local
and international governments have given to equity in the context of their
mobility agendas. A group of local mobility providers, used as cases, made
possible the application of the framework proposed in this research.
ever, depending on how long commuters remain away of public transit.
As cities hope to combat the negative externalities of motorized traffic
(e.g. congestion, emissions, accidents, etc) by offering safe alternatives
to transit through policies, plans and time-sensitive interventions (e.g.
interim bike paths), there are opportunities for decision-makers and
private mobility providers to align their priorities. In this context, this
thesis is placed right in between both sides, trying to encourage the use
of micro-mobility through the improvement of the provided services
and their monitoring to well serve all potential users.
F
igure
2:
average
trip
duration
versus
transit
ridership
during
covid
-19. s
ource
:
bay
wheels
.
0 250 500 750 1000 1250 1500 1750 2000 2250 2500 2750 -150% -100% -50% 0% 50%
01
/1
2/
19
08
/1
2/
19
15
/1
2/
19
22
/1
2/
19
29
/1
2/
19
05
/0
1/
20
12
/0
1/
20
19
/0
1/
20
26
/0
1/
20
02
/0
2/
20
09
/0
2/
20
16
/0
2/
20
23
/0
2/
20
01
/0
3/
20
08
/0
3/
20
15
/0
3/
20
22
/0
3/
20
29
/0
3/
20
05
/0
4/
20
12
/0
4/
20
19
/0
4/
20
26
/0
4/
20
03
/0
5/
20
10
/0
5/
20
17
/0
5/
20
24
/0
5/
20
31
/0
5/
20
07
/0
6/
20
14
/0
6/
20
21
/0
6/
20
28
/0
6/
20
Mean T
rip Dur
ation [sec] [Ba
y Wheels]
Demand f
or T
ransit [T
ransitApp]
Date [2020]
Percent Change Transit Ridership [TransitApp]
Percent Change Driven Km by Day [Waze]
Trip Duration [sec] (Bay Wheels)
1.2.2.
Scientific Relevance
This work contributes to existing scientific research related to smart
mobility, with specific focus on private mobility providers and open
data, as it suggests a systematic approach to assess the performance
of their service, broadening the scope from the operational level to
more strategic levels. As it proposes a framework, which is not applied
to a large number of cases due to scope limitations, it opens space for
future related research, not only by other scholars, but by consultants
interested in assisting municipalities on how to manage new mobility
services that are progressively occupying the public space. Within the
suggested framework, a number of literature-based indicators related
to certain policy goals can be assessed and adapted to the context of
private mobility services, which can also be further studied, criticized and
improved in future studies.
The following chapter (2) provides the theoretical basis prior the application
of the framework, exploring existing data standards for mobility and
their usefulness for monitoring policies via multi-criteria methods and
indicators. Chapter 3 regards at the methodological steps necessary to
respond the research questions, from data collection to the application
of the framework and its conceptual validation. Chapter 4 provides the
outcomes of the steps proposed in 3, using a group of cases. Chapter 5
and 6, respectively, are composed by final conclusions and references.
2.
theoretical
framework
Assuming that in the near future, the availability of accurate and
up-to-date mobility data will not be an issue for many cities in the world, the
demand for systematic frameworks able to deal with this data through
performance measures will increase. In this context, relevant questions
can arise in relation to the kind of data that is being exchanged and
manner it is structured; about the main institutional actors producing
or consuming those data; and the systematic approaches that can be
used to monitor urban mobility and the contexts in which they have been
applied. The following topics explore the exposed matters.
According to McKenzie (2020), a number of public transport agencies
nowadays are restricting certain types of occupation patterns in cities
and demanding that the crescent number of mobility services should
publicly share their data in order to be granted operational permits.
In this sense, other municipalities have quickly followed suit, which
have progressively straightened the communication gaps between
public transport authorities and private parties through open mobility
data specifications and their embedded open access APIs (Application
Programming Interfaces). Public initiatives such as the Mobility Data
Specifications (MDS), inspired by projects like General Transit Feed
Specification (GTFS) and General Bicycle Feed Specification (GBFS), aim
to provide standardized ways for regulatory bodies to ingest, compare
and analyse data from service providers, and to give them the ability
to express regulation in machine-readable formats. Similar, but more
limited aspects, can be expected from the other two mentioned formats
(GTFS and GBFS), since they are not able to take the message from
municipalities, but are quite valuable in showing travel patterns at the
user scale.
Data standards are also being used to enable the communication between
companies towards the diffusion of multimodal trips and provision of
a seamless experience to users. The concept of Mobility as a Service
(MaaS), conceived in Finland, aims to offer the opportunity for seamless
travel using any available transport mode, by integrating services from
data consumers, including journey planning, booking, ticketing, payments
(Sakai, 2019) and client complaints. The following figure illustrates a
2.1.
l
iterature
review
and
theoretical
Framework
basic functional structure of MaaS, in which a MaaS provider acts as
intermediary between the users and the transport providers, bundling
them to provide a variety of transport services (Sakai, 2019). In order
to facilitate the success of MaaS, transport operators are required
to standardize the digital form to facilitate access to their information
(Rijkswaterstaat, 2019).
In the last 10 to 15 years, a number of APIs have been developed and made
available in the transportation ecosystem by startups and public parties
in order to, either fulfill the gap of data availability and accuracy, or to
walk towards integrated and seamless mobility systems, thus finding new
market niches. The following figure illustrates a non-exhaustive overview
of available commercial and non-commercial APIs on the market to be
consumed by the general public and public or private sectors. Formats
such as GTFS (developed by google in 2006) and GBFS (developed under
the leadership of the North American Bikeshare Association) are currently
adopted by hundreds of companies and agencies as their data standard
to report on supply and demand transport systems (Figure 4).
F
igure
3:
standard
-
based
api
s
For
transport
operators
to
/
From
m
aa
s
providers
. s
ource
:
adapted
From
(Real-time) General Transit Feed Specification:
Developed under the leadership of Google and Trimet (a transit authority
in Oregon) in 2005, the General Transit Feed Specification (GTFS) is a data
specification that allows public transit agencies to publish their transit
data in a format that can be consumed by a wide variety of software
applications. GTFS stands out because it was conceived to meet practical
needs in communicating service information to passengers, not as an
exhaustive vocabulary for managing operational details. It is designed
to be relatively simple to create and read for both people and machines.
Even organizations that work with highly detailed data internally using
standards like NeTEx find GTFS useful as a way to publish data for wider
consumption in consumer applications (GTFS.org, 2020). The following
features are provided once a GTFS API is accessed: agency information;
stops, where vehicles pick-up and drop-off riders; transit routes; trips for
each route; stop times; calendar dates; fare attributes and rules; among
others. As the following figure shows, in relative terms, US and Europe
lead the production of GTFS data world-wide (transitfeeds.com, 2020).
2.3.
diving
deeper
into
mobility
data
standards
F
igure
4:
non
-
exhaustive
storyline
oF
commercial
and
non
-
commercial
mobility
api
s
. s
ource
: o
wn
production
.
F
igure
5: gtFs
data
Feeds
published
worldwide
. s
ource
: t
ransitFeeds
.
com
.
GTFS can be split into a static component, that contains schedule, fare,
and geographic transit information; and a real-time component, that
contains arrival predictions, vehicle positions and service advisories. Due
to it’s complexity, the feed is composed of a series of text files collected in
a ZIP file. Each file models a particular aspect of transit information. The
following figure represents a typical structure of a GTFS feed:
agency
fare attributes
fare rules
shapes
calendar
calendar dates
frequencies
transfers
feed_info
routes
trips
stop times
stops
General Bike Feed Specification:
Under the NABSA’s (North American Bikeshare Association) leadership,
the General Bikeshare Feed Specification (GBFS) was developed by a
cross-sector team of bikeshare system owners and operators, application
developers, and technology vendors. More than 200 bikeshare and scooter
systems worldwide have adopted GBFS since its release in November
2015 (Figure 7). The majority of bodies, either private or public, making
GBFS available are concentrated in US and Europe.
The specification has been designed with the following purposes (NABSA,
2020): provide real-time status of the system, do not provide information
whose primary purpose is historical, the data in the specification is
intended for consumption by clients intending to provide real-time (or
semi-real-time) transit advice and is designed as such. The following
information is provided once a GBFS API is accessed: details about
the system operator, location, year implemented; list of all stations,
their capacities and location; number of available bikes and docks at
each station and availability; bikes that are available for rent; hours of
operation, etc.
As GTFS, GBFS is also composed by a series of datasets that can be
connected to one another. The following are the most commonly found
elements in a GBFS API, but do not exclude the other files described by
the official specifications (NABSA, 2020):
- system_information > Describes the system including System operator,
System location, year implemented, URLs, contact info, time zone;
- vehicle_types > Array that contains the physical characteristics of the active
vehicles, including their form (e.g. bicycle, car, moped, scooter), propulsion
type (e.g. human, electric, combustion, etc), furthest distance in meters that
the vehicle can travel without recharging/ refueling, etc;
- station_information > Array of all stations that are considered public (e.g.
can be shown on a map for public use), including their location, reference points,
rental methods, area, capacity, etc;
- station_status > Includes time-sensitive information about the number
of available vehicles and docs, if whether the station is renting or returning
vehicles, etc;
- free_bike_status > Array containing data about vehicles that are currently
stopped, and whether reserved or not, disabled or not, etc;
- system_hours > array that provides the system hours of operation, including
calendar days, start and end times, etc;
- system_regions > Describes the regions the system is broken up into;
- geofencing_zones > Defines geofencing zones available in the system and
their link to physical stations if any;
Mobility Data Specification:
The Mobility Data Specification (MDS), a project of the Open Mobility
Foundation (OMF), is a set of Application Programming Interfaces
(APIs) focused on dockless e-scooters, bicycles, mopeds and carshare.
Inspired by projects like GTFS and GBFS, the goals of MDS are to provide
a standardized way for municipalities or other regulatory agencies to
ingest, compare and analyze data from mobility service providers, and to
give municipalities the ability to express regulation in machine-readable
formats (LADOT, 2018). According to Esri (2020), it defines three distinct
main components: providers (information to be consumed by regulatory
agencies); agency (to be consumed by mobility providers) and policy
(local rules that may affect the operation of mobility services). More
than 80 cities and public agencies around the world use MDS, and it has
been implemented by great number of mobility providers (OMF, 2020).
Their structure and openness are dependent on the context they were
developed.
Some examples of how cities can use MDS in practice can be mentioned:
verify how many vehicles are operating; verify whether vehicles are
being deployed equitably across neighbourhoods, determine whether
scooters are dropped off outside of a service area; inform future capital
investments such as dockless vehicle drop zones or urban furniture
zones; inform infrastructure planning efforts such as athe addition of
bike lanes; inform micro-mobility policy making; among others.
Planning sustainable and efficient transport systems is crucial for
reducing their negative impacts on the environment (Damidavicius et
al., 2019) and building more accessible cities. In order to facilitate such
changes, strategic and integrated planning approaches are necessary.
The ideas and concepts of sustainability and efficieny need to be
given operational forms if they are to influence current governance of
transportation systems. In this sense, systematic monitoring and
evaluations can provide information on the progress of the planning
and implementation processes, and the design or expected impact
of existing mobility measures. Regular monitoring and evaluation of
impacts to structure learning and improvement processes are seen as
best-practices to achieve sustainable mobility (Rupprecht Consult, 2019).
As a result, monitoring and evaluation indicators are acknowledged by
many policy makers and scholars as fundamental to follow the progress
of planning and implementing mobility plans, and their respective policy
goals; following realisation processes, appropriate decisions on further
measures should be adopted, taking into account monitoring data.
The mentioned monitoring systems adopt different methodologies to
“frame” indicators, organizing them in larger groups of criteria or factors
(e.g. composite or weighted indexes, policy goals, etc). According to
Gudmundsson et al. (2016), indicators of different scopes are usually
bundled together and linked to other information to serve as overall
assessment purposes. When indicators are used in an organizational,
goal-oriented setting, they can serve as performance measures in the
review or accounting application of plans. Their choice depends on the
2.4.
monitoring
urban
mobility
:
Frameworks
and
indicators
F
igure
8:
key
dimensions
in
building
an
indicator
Framework
.
t
able
1:
Framing
questions
and
answers
For
diFFerent
planning
Frameworks
. s
ource
: g
udmundsson
et
al
. (2016).
intended goals of governments (e.g. sustainability, accessibility, fairness,
among others), therefore must be organized and “framed”. Frameworks
are key in organizing and conceptualizing information and actions to
inform the development of desired outcomes.
Frameworks of all kinds are relevant as they suggest certain ways to
think, organize, measure and act. Their main advantage is that they can
provide structured ways to deal with communication needs: What should
be measured, how and to whom should the results be reported? At the
same time, this can be their limitation, since not all stakeholders might
agree with the proposed assumptions in the framework. One example is
the use of economic methods to evaluate impacts of transportation (Sager
and Ravlum, 2005; Bakker et al., 2010). No framing, however, is rarely a
good option. Frameworks bring focus, purpose, direction and attention
to planning processes. Three basic questions can be respondend to help
build indicator frameworks (see typical examples of frameworks providing
a different set of answers in Table 1) (Gudmundsson et al., 2016):
(a) “Why” is the information needed? refers to the intention and
application;
(b) “What” information is needed? refers to the specific impacts
measured; and
(c) “How” is the information delivered? refers the operation of the
proposed frameworks.
Gudmundsson et al., (2016) calls attention to key aspects on building a
framework. The ideal situation is represented in the following figure, in
which there is a mutual alignment between key dimensions involved in
the construction of indicators frameworks (intentions, procedures and
theory), but also highlights the risks of unbalanced situations, where
one concept dominates the formulation of the indicators. For instance,
concept-driven indicators often stem from scientists using theories not
well recognized by policy makers or the general public. In this sense,
the goal in building a monitoring framework would be to have a clear
understanding of the three dimensions and ensure an adequate balance
between them.
F
igure
9:
key
dimensions
in
building
an
indicator
Framework
. s
ource
:
adapted
From
g
udmundsson
et
al
. (2016).
In regards to the necessary qualities a planning framework should have
in order to be an effective promoter of sustainable development, scholars
at the Georgia Institute of Technology have proposed a set of attributes
that combine concerns for how to represent sustainability:
1. Comprehensiveness: Essential to ensure adequate and operational
representation of sustainability principles. It is believed that the
integration with a holistic view across the different pillars and across the
dimension of past and future needs is important for measuring whether
a development is likely to be sustainable or not.
2. Connection to strategic goals: Connecting measurement to goals and
objectives of a city or agency is believed to be essential for frameworks
that are embedded in governmental institutions. Otherwise there is a
risk that some objectives will become sidelined within an organization
in relation to others (e.g. performance of road traffic instead of equity).
3. Internal Integration: Frameworks should be integrated vertically and
horizontally through agencies to allow more effective management. This
refers primarily to large public and private organizations charged with
the development and management of major transportation systems.
4. Interactions: Frameworks shall provide the capability of capturing the
effects of interactions between variables. Does the framework describle
possible linkages between the impacts of interventions and sustainability
outcomes? Does it allow for the identification of synergies among impacts
to point toward most effective measures to enhance the analysed system?
(Pei et al., 2010)
5. Stakeholder perspectives: Urban mobility involves all participants in
society, which are organized as stakeholder groups. The active engagement
of different stakeholders in decision0-making is a fundamental principle
for sustainable development.
6. Agency capabilities and constraints: It is important to put particular
emphasis on what an agency can or is allowed to influence (Jeon and
Amekudzi, 2005). For each element within a transport system (e.g. roads,
rail, urban transport), there might be official documentation describing
the scope and purpose of their work and power to regulate the players
(e.g. mobility providers, users).
7. Flexibility and learning: Frameworks need to support the continuous
development of organizations, programs and policies over time. There is
a need to evolve and accumulate knowledge and evidence about causal
relations, sustainability impacts and effectiveness of measures.
Indicators frameworks are often structured and organized by
multi-criteria techniques. Urban mobility plans world-wide use multiple
criteria (and sub-criteria) to assess present and future scenarios to guide
spending on existing and future transportation projects according to
predefined policy goals or recommendations, and to systematically learn
from previous experiences, adjusting and improving planning activities
(Rupprecht Consult , 2019). The European Platform on Sustainable
F
igure
10:
typical
structure
oF
ahp. s
ource
:
adapted
From
s
aaty
(1980).
Among the various MCDA tools used to improve mobility systems, the
most commonly used is the Analytical Hierarchy Process (AHP), developed
and proposed by Saaty (Kramar et al. 2019). This might be due to its
accuracy, simplicity, and theoretically robust capacity for handling both
numerical and non-numerical measurements, and its ability to embrace
real-world factors in the model. Figure 10 illustrates the main elements
of an Analytical Hierarchy Process.
Within the AHP approach, further exaplained in chapter 3,
decision-makers make different judgements to multiple criteria, which when
combined with the measured indicators, can make the results more
reliable, closely relating them to local contexts and reduce potential
decision errors (Taleai and Amiri, 2017).
Urban Mobility Plans, in their methodology, proposes the identification
of strategic indicators and targets that allows monitoring progress made
towards realizing all objectives of a plan.
According to Boisjoly (2017), clear multi-criteria analyses, using well
defined indicators, provide greater transparency and typically foster
the inclusion of relevant aspects in decision-making processes (e.g.
accessibility, equality, sustainability). Multi-criteria decision analysis
(MCDA) have been recently used as a way to solve complex planning
problems involving more criteria and more decisions-makers. In this
sense, MCDA is regarded as an appropriate framework to be further
explored in this research, since it decreases uncertainty and improves
the quality of decisions.
The combination of theories explored in this thesis so far takes the
following stand-points as theoretical pillars to feed the next steps:
1) Despite various initiatives promoting “smartness” in urban transport
systems with strong participation and proactiveness of private actors,
little is still known about how these systems are performing in relation
to what cities might be aiming for in terms of urban planning and design
(Debnath, 2014). It is however essential that agencies are able to capture
the effects of changes in mobility systems, the interactions between
players and the impacts of mobility to point towards effective measures;
2) As mobility options proliferate, policy frameworks to govern mobility
through standardized and real-time data sharing are overdue. Such
frameworks should meet agencies’ capabilities, consider different
stakeholders’ perspectives (Pei et al., 2010) to give more proactiveness
to the public sector, enabling decision-makers to learn and keep up with
the pace that the market and trends in mobility are changing, instead of
being only reactive to them;
3) It is progressive the increase of awareness of different actors (public
and private) regarding the usefulness of open/ standardized mobility
data to help support research, public policy development, contribute to
public agency enforcement, operational management and transportation
planning. Such an ecosystem also foments the emergence of innovative
business models;
In a desirable urban scenario, the interactions between the public
sector and private services are illustrated in the following feedback loop
suggested in Figure 11. After requesting and receiving (2) real-time
standardized data from the mobility ecosystem (interactions between
different stakeholders, 1), municipalities would be able to assess it, and
measure the extent that the services are compliant to existing policy goals
(3), either strategic and/ or operational, using predefined systematic
frameworks. In the end of the cycle, data-based policies would feed the
mobility ecosystem back again (4).
2.5.
Framing
policy
goals
and
compliance
oF
private
keywords: seamless
mobility systems,
smart mobility
solutions, openness
and standardization of
mobility data
keywords: monitoring
urban mobility,
evaluation frameworks
performance indicators,
multi-criteria
3.
research
design
Research design influences the validity and trustworthness of the
research results. Its outcomes could be used for policy makers in real
planning practices and, therefore, insufficient or invalid research designs
could potentially create erroneous applications. The proposed approach
is perfomed from a Post-Positivist perspective. Such philosophy, unlike
positivists, deny the total independence and objectivity of the researcher
and the object, while opening space for qualitative research. In this work,
a quantitative dominant mixed methods approach relies on a quantitative
(e.g. indicators) post-positivist view of the research process, where the
addition of qualitative data (e.g. experts opinions) is likely to benefit most
research projects (Johnson and Turner, 2007). The main justification for
this methodological choice is that a single dataset type (quantitative x
qualitative) would not be able to answer all research questions. Because
the performance of mobility providers should not only be measured
quantitatively, but should be fed with experts’ perceptions depending on
the urban context, a mixed-methods design is applied where the data
gathering is partially quantitative and qualitative. In the following topics,
the methodological approach is proposed to respond to the objectives of
this research.
The approach can be divided in 5 major phases, further explored in topic 3.3:
(1) The proposal of a monitoring framework, including its accommodation
in real-world policy scenarios (case studies), where the quality of private
mobility services would be weighted and benchmarked using a
multi-criteria tools, and quantitative and qualitative performance measures;
(2) Data collection, from the requests for mobility data through the use
of open APIs to their treatment and processing for future assessments;
(3) Exploratory data analysis, including the application of basic statistical
measures (e.g. averages); and production of maps and charts (e.g.
histograms); (4) Compliance assessments, in which the indicators are
normalized and aggregated to compose single measures; (5) Framework
evaluation, where matters related to the practical validity and reliability
of the application in case studies will be addressed qualitatively. The
following figure illustrates the proposed methodological steps:
3.1.
r
esearch
strategy
3.2.
a
pplication
oF
the
Framework
in
a
case
study
In order to investigate the applicability of the proposed framework in
real urban settings and explore possible indicators regarding existing
policies, the utilization of case studies can be considered relevant (van
Thiel, 2014). The quality of the results will depend on the chosen scope
and the criteria used to delimit it. In this sense, the scope of the case
studies is defined as: the data standard(s) to be obtained via mobility APIs
and analyzed; the private service(s) to be monitored; the transportation
policies and locations to serve as real-world urban scenarios for the study.
It is important to outline the interdependence between the factors of the
scope - defined the framework, its application only has relevance and
validity if there are up-to-date mobility data made available by transport
operators, as the city or region has reasonably recent urban transportation
plans/ visions with clear goals that can also be addressed by private
providers. In addition, the following criteria are being considered to guide
the delimitation of the scope (Joumard et al. , 2011): a) representation:
validity, reliability, sensitivity; b) operation: measurability, data availability,
ethical concerns; c) policy application: transparency, interpretability,
target relevance, actigin ability. The mentioned criteria will be further
explained in further stages of the study.
As mentioned, emergent modes are being regarded as the future
of transportation in cities around the world, with potential to create
more sustainable travel behaviours since they can be very attractive
alternatives for the “first- and last-mile” commuting trips, or can also be
associated with physical and leisure activities. In this context, a number of
companies started offering micro-mobility services and using the public
space of cities. In parallel, joint programs involving private providers
and public authorities were initiated to provide last-mile services (e.g.
Lyft Bay Wheels, Metro Bike Share). Along with the availability of
micro-mobility, data becomes also more easily available, with cities demanding
its openness to private providers in exchange for operation licences.
Figure 13 shows a non-exhaustive distribution of micro-mobility service
providers (private, public or jointly initiated) according to the quantity of
The framework itself can be regarded as predominantly quantitative, since
mathematical operations shall be used to measure, weight and combine
the performance indicators, which can be qualitatively interpreted, but
shall also be quantitatively operationalized. The collection, processing
and analysis of mobility data imply the use of statistical tools to deal
with relatively large amounts of points and variables. Evaluating the
applied framework involves the interpretation of the indicators by key
stakeholders from public authorities from the case studies, in order to
learn from them possible research drawbacks and to open windows for
future improvements.
0 20 40 60 80 100 120 United States Germany India Spain China France United Kingdom Brazil Singapore Sweden Canada Italy Austria Israel Netherlands Australia Colombia Denmark Mexico Norway Poland United Arab Emirates Chile Ireland Russian Federation
Quantity
C
ount
ry
Last-mile mobility service providers and GBFS APIs available
GBFS providers Private providers R² = 0.87948 0 20 40 60 80 100 120 140 0 20 40 60 80 GB FS da ta pr ov ide rs Private providers
