ISSN 1055- 1425
February 2006
This work was performed as part of the California PATH Program of the
University of California, in cooperation with the State of California Business,
Transportation, and Housing Agency, Department of Transportation, and the
United States Department of Transportation, Federal Highway Administration.
The contents of this report reflect the views of the authors who are responsible
for the facts and the accuracy of the data presented herein. The contents do not
necessarily reflect the official views or policies of the State of California. This
report does not constitute a standard, specification, or regulation.
Final Report for Task Order 5103
CALIFORNIA PATH PROGRAM
INSTITUTE OF TRANSPORTATION STUDIES
UNIVERSITY OF CALIFORNIA, BERKELEY
Enhanced Transit Strategies: Bus Lanes with
Intermittent Priority and ITS Technology
Architectures for TOD Enhancement
UCB- ITS- PRR- 2006- 2
California PATH Research Report
Michael Todd, Matthew Barth,
Michael Eichler, Carlos Daganzo,
Susan A. Shaheen
CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS
Enhanced Transit Strategies: Bus Lanes with
Intermittent Priority and ITS Technology Architectures
for TOD Enhancement
California PATH MOU 5103 Final Report
Michael Todd, Matthew Barth
College of Engineering- Center for Environmental Research and Technology
University of California, Riverside
Michael Eichler, Carlos Daganzo
Department of Civil and Environmental Engineering
University of California, Berkeley
Susan Shaheen
California Partners for Advanced Transit and Highways ( PATH)
University of California, Berkeley
PATH Research Report: Enhanced Transit Strategies
Abstract
Due to increases in congestion, transportation costs, and associated environmental impacts, a
variety of new enhanced transit strategies are being investigated worldwide. The transit- oriented
development ( TOD) concept is a key area where several enhanced transit strategies can be
implemented. TODs integrate transit, residential, retail and/ or commercial entities into a
compact, pedestrian- friendly community, thereby reducing private car usage and increasing
transit use. This research report addresses two enhanced strategies within the TOD framework:
1) using Bus Lanes with Intermittent Priorities ( BLIPs) to enhance bus transit; and 2) addressing
how and what Intelligent Transportation System ( ITS) technology can be used within TOD
system architectures. With respect to 1), it has been shown that the implementation of BLIPs for
bus rapid transit can greatly increase system efficiencies without compromising the level of
service for other facility users. The basic analysis in this report shows that both conservative and
liberal approaches have similar impacts to traffic and identical benefits. The macroscopic
analysis illustrates that traffic disturbances caused by BLIP activation will not slow down
subsequent buses, and that roads with medium traffic demand can easily support a BLIP
implementation. The microscopic analysis provides some quantitative equations that can help
decision makers determine whether a given intersection can be outfitted with a BLIP
implementation within predefined parameters. A framework for cost- benefit analysis was
provided for BLIP implementation. With respect to 2), it has been shown that transportation
efficiency and effectiveness within a TOD can certainly be enhanced with the integration of ITS
technology. This project report has identified technology bundles and architectures that have the
greatest potential for increasing mobility. Further, it has demonstrated that ITS technologies
implemented in a well- integrated fashion will promote transit efficiency and convenience and
lead to transit usage beyond levels currently observed.
i
PATH Research Report: Enhanced Transit Strategies
Summary
Increases in congestion, transportation costs, and associated environmental impacts continue to
promote the research, planning, and development of enhanced transit strategies. The transit-oriented
development ( TOD) concept, often synonymous with “ transit village”, integrates transit,
residential, retail and/ or commercial entities into a compact, pedestrian- friendly community. The
ultimate transportation objective relative to a TOD is to reduce private car usage with an
associated increase in transit ridership. Previous research indicates that residents living in
developments near stations are five to six times more likely to commute via transit than other
residents in a region. Additionally, a proportional relationship has been found between urban
density and transit use. Relative to bus transit, ridership associated with buses can promote
greater efficiencies through the implementation of bus rapid transit ( BRT) strategies. One of the
most promising areas of BRT enhancements is Bus Lanes with Intermittent Priorities ( BLIPs).
Successfully transitioning individuals from private vehicle usage to transit ridership is a complex
transition involving an array of socio- economic variables. This study focuses on strategies, such
as BLIPs, and ITS implementation architectures within TODs to promote the adoption of transit.
Private vehicle users must perceive significant benefits for adopting transit in place of a personal
vehicle. These often include: economic, time, convenience, or environmental benefits. While the
environmental, economic, and travel- time benefits associated with transit are quantifiable, user
convenience and system efficiency are more variable due to user perception. With other factors
being fairly equivalent, individuals will choose the transportation option with the most consistent
convenience. Those who own automobiles will compare the convenience of private vehicle use
relative to the level of convenience obtained through transit. Significant transit improvements are
desirable in the area of perceived convenience and associated time savings. Enhanced transit
strategies are continuing to expand mobility options, system efficiencies, and level of
convenience associated with transit. These improvements are increasingly being achieved
through a variety of ITS implementations.
Ever expanding ITS technology improvements related to communications and electronics
continue to create exciting options for TODs and BRT. These improvements include BLIPs,
smart parking, electronic payment services, innovative mobility modes, enhanced traffic
management, vehicle monitoring and control, carsharing, and driver and traveler services. The
following enhanced strategies have been evaluated:
Bus Lanes with Intermittent Priorities— uses changeable message signs, traffic signal
priority, automatic vehicle location, and in- pavement lights to yield right- of- way to the
bus. Such a system would ultimately decrease route travel times and increase system
reliability by ensuring schedule adherence. Ideal system configurations and operational
methods are discussed.
Implementation of TOD System Architectures— this analysis focuses on: 1) the
integration of compatible ITS strategies into an open architecture structure; 2) the
evaluation of suitable TOD architectures that integrate common ITS components into a
single modular networked system; 3) the integration of specific advanced personal
vehicle services within a TOD for improved personal mobility; and 4) modular synthesis
i i
PATH Research Report: Enhanced Transit Strategies
of advanced personal vehicle services ( APVS) into a TOD environment including: low-speed
modes, carsharing, smart parking, and Elockers.
As transportation networks become overburdened with increases in travel demand, system
efficiency requirements must also increase to maintain an acceptable level of service.
Transportation systems are continually being augmented with ITS technologies to maintain these
needed system efficiencies. The analysis of BLIPs has proven the system effectiveness
associated with combining specific ITS strategies with BRT scenarios. A well- integrated BRT
system utilizing a BLIP configuration can increase bus transit efficiencies, while not
compromising the level of service for other users of the facility.
The modular synthesis of multiple ITS strategies into a networked system requires the strategic
development of individual technology components. Through the exploration of multiple TOD
architecture scenarios for advanced personalized vehicle services, the inter- system
communication techniques and compatibility becomes the foremost issue. Utilizing an open
architecture Internet- based backbone allows for individual ITS components to be melded into a
single TOD servicing system. Users of the TOD perceive and access a single system to service
all their associated transit needs.
This study presents new and exciting ITS technology solutions for enhancing transit
deployments. The ITS strategies have demonstrated the potential to provide transit users with
increased mobility while limiting the dependence on the private vehicle. It has been shown that
transportation efficiency and effectiveness within a TOD can certainly be enhanced with ITS
synthesis. Additionally, implementation of BLIPs for BRT can greatly increase system
efficiencies without compromising the level of service for other facility users. The goal of this
report has been to identify technology bundles and architectures that have the greatest potential
for increasing mobility. This study has demonstrated that ITS technologies implemented in a
well- integrated fashion will promote transit efficiency and convenience and lead to transit usage
beyond levels currently observed.
ii i
PATH Research Report: Enhanced Transit Strategies
Contents
1. INTRODUCTION................................................................................................................... 1
2. BACKGROUND..................................................................................................................... 3
2.1. INNOVATIVE TRANSIT BUS CONCEPTS............................................................................... 3
2.2. TRANSIT- ORIENTED DEVELOPMENT ( TOD) AND COMPONENTS....................................... 4
2.2.1. Shared- Use Vehicle Systems........................................................................................ 5
2.2.2. Smart Parking Management ......................................................................................... 6
2.2.3. Low- Speed Transportation Modes ............................................................................... 8
3. ITS TECHNOLOGY AND IMPLEMENTATIONS ............................................................. 11
3.1. NATIONAL STANDARDS AND GUIDELINES ....................................................................... 11
3.2. SPECIFIC TECHNOLOGY..................................................................................................... 12
3.2.1. Wireless Communications.......................................................................................... 12
3.2.2. Resource Access Control............................................................................................ 15
3.2.3. Trip and/ or Resource Performance Data Acquisition ................................................ 18
3.2.4. Navigation Systems and Automated Vehicle Location Capability ............................ 18
3.2.5. System Messaging...................................................................................................... 20
3.2.6. System Management .................................................................................................. 20
3.2.7. Reservation Management ........................................................................................... 20
3.2.8. Accounting Systems ................................................................................................... 21
3.3. IMPLEMENTATIONS............................................................................................................ 22
3.3.1. Carsharing .................................................................................................................. 22
3.3.2. Station Cars ................................................................................................................ 23
3.3.3. Other Shared- use Vehicle System Models................................................................. 23
4. INTELLIGENT BUS PRIORITY LANE ANALYSIS.......................................................... 25
4.1. BASIC ANALYSIS ............................................................................................................... 26
4.1.1. Scenario Description .................................................................................................. 26
4.1.2. Supporting Concepts .................................................................................................. 26
4.1.3. Overview of Approaches............................................................................................ 28
4.1.4. General Findings ........................................................................................................ 29
4.1.5. Macroscopic Analysis ................................................................................................ 31
4.2. DETAILED ANALYSIS......................................................................................................... 36
4.2.1. Analysis Overview ..................................................................................................... 36
4.2.2. Supporting Concepts .................................................................................................. 36
4.2.3. Other Factors .............................................................................................................. 40
4.2.4. Feasibility Analysis .................................................................................................... 40
4.2.5. Benefit Analysis ......................................................................................................... 45
4.2.6. Reduced Travel Time Variation ................................................................................. 52
4.2.7. Qualitative Benefits.................................................................................................... 53
4.3. COST ANALYSIS................................................................................................................. 54
4.3.1. Increased travel time for traffic .................................................................................. 54
4.3.2. Installation and operating costs .................................................................................. 54
4.4. BENEFIT/ COST COMPARISON............................................................................................ 54
5. TOD SYSTEM ARCHITECTURE ANALYSIS................................................................... 56
5.1. MODULAR ITS IMPLEMENTATION FOR TOD.................................................................... 57
5.2. REVIEW OF SYSTEM ARCHITECTURE SCENARIOS ............................................................ 57
iv
PATH Research Report: Enhanced Transit Strategies
5.2.1. Physical Network Characteristics............................................................................... 58
5.2.2. Communication Protocol............................................................................................ 59
5.2.3. Ancillary Communications......................................................................................... 60
5.2.4. General ITS Micro- architecture for TOD Enhancement............................................ 60
5.3. PROPOSED DESIGN FOR PLEASANT HILL TOD................................................................. 63
5.3.1. Intermediate- Level Design......................................................................................... 64
5.3.2. Advanced Design ....................................................................................................... 66
5.3.3. Distributed Database with Distributed Server Configuration .................................... 67
5.4. COST EFFECTIVENESS ANALYSIS FOR VARIOUS ARCHITECTURES .................................. 69
5.4.1. Development Costs..................................................................................................... 69
5.4.2. Implementation Costs................................................................................................. 69
5.4.3. Operational Effectiveness........................................................................................... 70
6. PROPOSED NEXT STEPS ................................................................................................... 72
6.1. INTELLIGENT BUS PRIORITY LANE................................................................................... 72
6.2. ITS IMPLEMENTATION FOR TOD ...................................................................................... 73
7. CONCLUSIONS AND FUTURE WORK............................................................................. 74
8. REFERENCES..................................................................................................................... . 76
APPENDIX A: LITERATURE REVIEW OF BUS LANE INTERMITTENT PRIORITY ........ 82
APPENDIX B: LITERATURE REVIEW OF ITS TECHNOLOGY FOR TODS ....................... 90
v
PATH Research Report: Enhanced Transit Strategies
1. Introduction
Travel demand in California continues to steadily increase, due primarily to California’s
expanding population growth. Most of California’s current travel demand is satisfied with the
automobile traveling on an expansive roadway system, often as single occupant vehicles.
However, the roadway system is no longer expanding with increased travel demand; as a result,
congestion has become a serious problem in terms of cost, safety, energy, and the environment.
It is clear that the state’s transportation system will need to provide for more efficient and
flexible mobility options beyond the standard use of automobiles. Transit can play a major role
in alleviating these problems; however, the majority of current transit systems are not very
flexible nor reliable. What is needed are innovative ideas that can provide integrated door- to-door
services to reduce travel times and increase ridership.
Transit agencies are seeking new ways to increase ridership and to provide better service with
limited resources. In recent years, many transit agencies have investigated several options,
including non- fixed- rail systems, such as Bus Rapid Transit ( BRT) as well as promoting Transit-
Oriented Developments ( TOD). In general, TODs promote transit use through the integration of
multiple transit options in high- density developments consisting of residential, commercial, and
retail entities. TODs have been demonstrated to increase transit usage, elevate the pedestrian
mode, and reduce private vehicle use [ Arrington, 2003].
Crucial to any transit option ( including BRT and TODs) is the use of Intelligent Transportation
System ( ITS) technologies. New technology expands the different transportation possibilities
and allows for significant improvements in mobility. Through the use of advanced transit modes,
innovative feeder options, integrated ITS technologies, shared- use vehicle systems ( i. e., short-term
vehicle rentals), and intelligent parking services, TODs have the potential to improve
personal mobility while enhancing the livability within a community. Further, BRT systems can
be enhanced through the application of ITS and new operational concepts, resulting in enhanced
mobility for bus riders in and around the transit community.
In 2004, this study was initiated to explore two key options which combine technological
advancements, operational improvements, and flexible approaches that could lower travel times,
enhance reliability, connectivity, and system appeal; and ultimately lead to increased transit
ridership. These two options include:
Bus Lanes with Intermittent Priority— this innovative concept employs changeable
message signs, traffic signal priority, automatic vehicle location, and in- pavement lights.
During hours of Intermittent Bus Priority operation, other vehicles can also make use of
the lanes. As a bus approaches, however, other vehicles are instructed to leave the lane,
yielding right- of- way to the bus. Other vehicles are instructed to maneuver to the other
lanes through a variety of methods: overhead and roadside signalization, in- pavement
lights, etc. Additionally, the bus would receive signal priority at intersections. Under
optimal conditions, the bus would only need to stop at bus stops, regardless of roadway
traffic conditions. Such a system would ultimately decrease route travel times and
increase system reliability by ensuring schedule adherence.
1
PATH Research Report: Enhanced Transit Strategies
Advanced Personalized Vehicle Services— these services combine several innovative
mobility solutions, including: shared- use vehicle systems, linkages to transit, small
electric vehicles, and advanced electronic and wireless communication devices. ITS
technology is used to facilitate reservations, billing, vehicle access, and traveler
information; as well as smart parking management services.
The results of this research are described in detail in this report. It is expected that these results
could take California closer to its goal of making transit a more competitive mobility option to
the single occupancy vehicle, particularly in congested corridors and regions, such as the Bay
Area. As part of this report, the two innovative mobility options are examined in terms of cost-effectiveness,
impacts on travel times, reliability and/ or flexibility, and how these strategies
could be bundled to offer greater benefits.
As part of this research program, specific project tasks were carried out:
Task 1: Detailed literature reviews were conducted on several topics, including ( but not
limited to) bus rapid transit, traffic signal priority, automatic vehicle location systems,
changeable message signs, in- pavement lights, shared- use vehicle systems, wireless
communications technology, and a variety of transportation system architectures;
Task 2: Based on the results of the literature review, a new Bus Lane with Intermittent Priority
( BLIP) concept was developed, as well as new system architectures that can be used
for integrated technology applicable to transit- oriented developments;
Task 3: These innovative concepts were then analyzed in detail, examining the cost-effectiveness
and the impacts on travel times, reliability, and flexibility; and
Task 4: Finally, proposed next steps were examined for these innovative options.
These research tasks were carried out by two research teams, one from UC Berkeley ( focusing
on bus lanes with intermittent priority) and the other from UC Riverside ( focusing on new
system architectures for TODs). In Chapter 2, the authors provide a brief overview of several
key transportation concepts to this study. Next, Chapter 3 describes the results of the technology
evaluation, describing different ITS components that are applicable to bus operations and TODs.
Chapter 4 then describes the BLIP concept in detail and provides a detailed analysis. Chapter 5
describes the integrated architectures for TOD development and associated analysis. In Chapter
6, the authors discuss proposed next steps with recommendations on how the technologies can be
implemented. Finally, Chapter 7 provides a summary and conclusions. The report also contains
two appendices, including detailed annotated bibliographies from the literature review.
2
PATH Research Report: Enhanced Transit Strategies
2. Background
This section of the report includes a brief overview of innovative transit bus concepts and
advanced personalized vehicle services.
2.1. INNOVATIVE TRANSIT BUS CONCEPTS
Buses that operate in mixed traffic lanes are subject to delays caused by traffic congestion,
reducing the appeal of bus transit. On the other hand, bus lanes provide excellent right- of- way to
transit vehicles. However, the reduction in private vehicle capacity of a traditional bus lane can
only be justified along roadways with very frequent or critical bus service, such as a BRT
system. As a compromise between dedicated bus lanes and buses operating in mixed traffic
lanes, the concept of Bus Lanes with Intermittent Priority or BLIPs can be implemented. With
BLIP, other traffic can make use of the lane as normal. However, as a bus approaches, other
vehicles are instructed to safely leave the lane ( or are prevented from entering the lane), yielding
the right- of- way to the bus. Dynamic signage can communicate the status of the BLIP to other
users of the roadway ( e. g., overhead signalization, roadside signalization, in- pavement lights,
etc.).
BRT systems often incorporate a variety of features to improve overall service. One feature that
is often employed is Transit Signal Priority ( TSP). In general, TSP can decrease bus travel times
by allowing buses to preempt or extend traffic signals to allow the transit vehicle to proceed
through an intersection. A handful of studies have documented the benefits of TSP
implementations, such as [ Balke et al, 2000; Banerjee, 2001; Cima et al, 2000; Duerr, 2000;
Furth et al, 2000; Garrow et al, 1998; Hunter- Zaworski et al, 1995; Janos et al, 2002; Kloos et al,
1995; Lin, 2002; Nash et al, 2001; and Skabardonis, 2000]. These and other references are cited
in the annotated bibliography in Appendix A.
Another option for BRT ( or other enhanced bus service) is the concept of an Intermittent Bus
Lane ( IBL): in this case, a lane is reserved for bus use, but it also allows private vehicle traffic to
use the lane when not in use by the bus. One study has proposed an IBL strategy: [ Viegas et al,
2001]. This IBL strategy never requests traffic to leave the lane to accommodate the bus; instead,
it restricts traffic from changing into the bus lane and relies on TSP to “ flush the queues” at
traffic signals. The BLIP concept proposed here is similar to this IBL concept; however, it clears
traffic out of the lane reserved for the bus when necessary, not relying on TSP. As a result, the
BLIP concept is easier and less expensive to implement.
The BLIP concept is also related to the idea of a queue jump lane [ Rosinbum et al, 1991; TRB,
2000; Mirabdal et al, 2002]. Widening the roadway near key intersections provides queue jump
lanes. These lanes only allow buses and right- turning vehicles to enter, enabling the bus to “ jump
the queue” of traffic at the signal. These lanes often have special signalization that allows the bus
to pull into the intersection before the vehicles in the other lanes, giving the bus priority as it
returns to the through- traffic lane. Unlike queue jump lanes, BLIPs require no additional right-of-
way and again should be less expensive to implement.
3
PATH Research Report: Enhanced Transit Strategies
2.2. TRANSIT- ORIENTED DEVELOPMENT ( TOD) AND COMPONENTS
Many transit organizations and communities are participating in the creation of commercial,
retail, and residential developments proximal to transit facilities. While a variety of
configurations and definitions can be found for a TOD, there is general consensus among transit
professionals that a TOD consists of “ a pattern of dense, diverse, pedestrian- friendly land uses
near transit nodes that, under the right conditions, translates into higher patronage” [ TCRP,
2004]. There are multiple types of transit- related developments that are often discussed in close
association with TODs. These include:
• Transit Adjacent Development ( TAD),
• Transit Village, and
• Transit Joint Development ( TJD).
TADs are proximal to a transit station and lack significant integration, while TJDs primarily
describe the development relationship between transit authorities, governmental bodies, and
business organizations [ Cervero, 2002]. Transit joint development is considered a sub- set of
transit- oriented development in which the development occurs on or adjacent to land owned by
the transit agency; the transit agency shares in some of the revenue generated by the project or
where there is some physical alteration made to the transit station as a result of the project
[ TCRP, 2004]. Developments that occur next to a transit location and do not fully integrate
transit into the development are often referred to as TADs. TADs often lack key pedestrian-friendly
components and are frequently smaller developments compared to TODs [ Arrington,
2003]. The TJD terminology is frequently used to describe multiple interests involved in the
development versus the transportation modes being promoted. Transit Village definitions are
generally synonymous with TOD definitions as discussed on a transit village- dedicated website
( www. transitvillages. org), as well as the Federal Highway Administration’s case study of the
Fruitvale Transit Village in Oakland California [ FHWA, 2005].
A review of TOD definitions has revealed some common similarities among most TOD
descriptions [ Cervero, 2002]. These include:
• Mixed- use development,
• Development that is close to and well served by transit, and
• Development that is conducive to transit ridership.
The potential success of transit is strongly correlated to how well the community design
promotes transit use. Mass transit designs inherently have significant distance between locations
( stations) where users can enter or exit the transit mode. This transit characteristic often requires
users to utilize another mode of transportation at either end of their transit- based trip leg. This
implies that an individual’s origin and/ or destination is often beyond the preferred walking
distance of a transit stop ( i. e., greater than one- quarter mile). This indicates that an overall transit
system must integrate effective mass transit services ( e. g., bus, bus rapid transit, train, subway,
4
PATH Research Report: Enhanced Transit Strategies
shuttle) as well as convenient feeder options ( e. g., bike access, taxi, low- speed vehicles, personal
vehicle parking, etc.).
TOD development is a complex process typically involving a multitude of stakeholders,
including: transit agencies, private developers, environmental groups, alternative transportation
advocates, residential developers, private retailers, and private transportation service providers.
Most interest groups agree that, if successful, TODs can yield many benefits, including increases
in transit ridership and profits to public and private partners [ TCRP, 2004]. The same TCRP
report states the top five transit agency motivations for engaging in TODs are:
1. Increasing ridership,
2. Promoting economic development,
3. Raising revenues,
4. Enhancing livability, and
5. Expanding housing choices.
In this section, a brief review is provided of several components that can play an important role
in transit- oriented developments. These include shared- use vehicle systems ( i. e., short- term
vehicle rentals, such as carsharing), smart parking management, and low- speed modes.
2.2.1. Shared- Use Vehicle Systems
There has been significant interest in shared- use vehicle systems over the last decade as an
innovative mobility alternative. The general principle of shared- use vehicle systems is that
individuals can access a fleet of shared vehicles ( ranging from cars to bikes and scooters) on an
as- needed basis, rather than using their personal vehicles for all trips. There are many potential
advantages of shared- use vehicle systems, including better vehicle use ( leading to higher
transportation efficiency), cost savings to the user, energy/ emissions benefits, and improved
access to established transit operations. For further information on the history and benefits of
shared- used vehicle system, see [ Shaheen et al., 1998; Britton et al., 2000].
Over the last several years, numerous shared- use vehicle services have developed that reflect
different operational models ( or market segments) and purposes. A classification system for
categorizing different shared- use vehicle system models, ranging from neighborhood carsharing
to station car systems ( i. e., shared vehicles directly linked to transit), was developed in 2002
[ Barth & Shaheen, 2002]. The predominant shared- use vehicle model is neighborhood
carsharing, where individuals in dense metropolitan areas access shared- use vehicles distributed
throughout neighborhood lots. Indeed, this is the prevailing approach in Europe and commercial
shared- use services in North America. Station car systems are another model, where vehicles are
closely linked to transit stations to enhance access. Station cars are often shared, although not
always. Some of the more innovative shared- use vehicle service providers today are combining
elements of both traditional carsharing and station cars, forming what are called “ hybrid” models
[ Barth & Shaheen, 2002]. As of July 2005, U. S. carsharing programs collectively claimed 76,420
members and operated 1,192 vehicles [ Shaheen et al., 2005a].
5
PATH Research Report: Enhanced Transit Strategies
When integrated within a TOD, shared- use vehicle systems can enhance mobility significantly.
The shared- use vehicle system can provide transit users with convenient personalized
transportation to their final destination. For many transit riders, the TOD transit station will
likely not provide the door- to- door convenience associated with privately owned vehicles. A
carefully designed shared- use vehicle system integrated within a TOD can get the transit user
closer to the personal mobility associated with private vehicle ownership. To achieve the
optimum level of convenience, well integrated ITS technologies should be integrated with the
shared vehicle system and corresponding transit modes.
One of the key elements of modern- day shared- use vehicle systems is the application of ITS
technologies. These technologies can enhance shared- use vehicle services by improving their
overall efficiency, user- friendliness, and operational manageability. Several ITS technology user
services [ U. S. DOT, 2005] can be applied: 1) dispatching and reservation systems so that users
can obtain system information, check- out vehicles, and make reservations over the web, by
phone, kiosk, etc.; 2) smartcard technology to assist with vehicle access control; 3) on- board
navigation and travel information to assist system users; and 4) intelligent communication and
tracking systems to provide vehicle location/ identification, emergency messaging, and electronic
debiting. Much of this advanced technology has been developed and applied in shared- use
vehicle research programs, such as the University of California- Riverside IntelliShare testbed
[ Barth et al., 2000] and the Carlink II program [ Shaheen et al., 2000].
Commercial carsharing organizations in North America have increasingly added technology to
their systems, where 70 percent of U. S. shared- use vehicle organizations have advanced
operations; 24 percent provide partially automated services; and six percent offer manual
services ( as of 2005, see [ Shaheen et al., 2005a]). In Canada, 73 percent of the carsharing
organizations have partial automation and 18 percent manual operations [ Shaheen et al., 2005a].
In Shaheen et al.’ s ( 2005a) technology analysis, manual operations include operator phone
services and in- vehicle trip logs; partially automated systems are automated reservations via
touch- tone telephone or Internet or both; and advanced operations involve smartcard access,
reservations, billing, automated vehicle location, and cellular/ radio frequency communications.
As shared- use vehicle systems continue to expand and multiply, the penetration of ITS
technology use will only increase as manually managing larger fleets and more diverse user
markets ( e. g., one- way trip rentals) becomes more difficult with increased scale.
The integration of shared- use vehicle systems into TODs has been slow to develop.
Nevertheless, emerging ITS technology developments are allowing shared- use vehicle systems
to be more feasible and economically viable for TOD integration.
2.2.2. Smart Parking Management
It is well known that parking is costly and limited in almost every major U. S. city, contributing
to increased congestion, air pollution, driver frustration, and safety problems. Furthermore,
limited parking can also constrain transit ridership in dense regions. As a potential solution to
many of these parking problems, smart parking management can be applied as an ITS solution
and is crucial for a TOD to succeed. Smart parking management is the use of advanced
technologies to help direct drivers efficiently to available parking spaces at transit stations ( and
other high- activity locations), encouraging transit ridership, lessening driver frustration, and
reducing congestion on highways and arterial streets. Smart parking approaches range from
6
PATH Research Report: Enhanced Transit Strategies
dynamic displays on roadway signs informing drivers of location and parking lot capacity, to
providing space availability, location, and pricing information through the Internet and/ or cell
phones.
In Europe, there are several smart parking systems that are replacing traditional paid parking
with real- time communications and payment systems via mobile phone. Recently, European
cities have integrated ITS technologies into intermodal transportation centers, such as transit
park- n- ride lots, to provide real- time information to motorists regarding availability and
electronic/ wireless parking payment services. This includes dynamic message signs ( DMS) and
changeable message signs ( CMS) that provide motorists real- time parking information ( see, e. g.,
[ Cervero, 1998]). According to a mobile company in Ireland, 70 percent of individuals in most
western European countries have mobile phones, and penetration rates increase among motorists.
Many European companies and municipalities use a smart paid parking platform that works on
normal mobile phones via the Internet. Advantages for customers include no coins/ exchange, a
lower chance of parking tickets, and a reduction in overfed meters ( as demonstrated in Easy Park
in Oslo, Norway; see www. easypark. net). In the U. S., intermodal transportation parking ( also
known as “ commuter lot parking”) began at gas stations along a Detroit transit line in the 1930s
[ Maccubbin & Hoel, 2002]. It is now common for cities and states to have transportation demand
management ( TDM) programs that include such commuter facilities or “ park- n- ride” lots to
better manage travel demand [ Maccubbin and Hoel, 2002]. Beyond park- n- ride lots and transit
station lots, little innovation has been attempted to better manage parking resources at critical
rail and bus lots in dense urban regions in U. S. cities. Only recently have researchers begun to
investigate smart parking management, such as smart parking field operational test linked to
transit. This test involves communication technologies to help manage existing parking spaces at
and around a BART station to increase space availability and transit access [ Shaheen et al.,
2005b].
As previously described, TODs must accommodate the individuals whom use a personal
automobile for some percentage of their transportation. Therefore, having an efficient and
convenient transition of individuals from their personal vehicle to transit is of utmost
importance. The transition of individuals from their personal vehicle to transit can be enhanced
through smart parking technology.
Many variations of smart parking have been implemented. These include autonomous parking
garages to reserved parking spaces. There are a few defining characteristics associated with
smart parking at a TOD:
1) The land area in and around the TOD is of high value, and therefore, some type of
efficient parking structure is nearly always a design preference;
2) The location upon which the driver departs from their vehicle needs to be within a
convenient distance to the transit stop or feeder service ( e. g., shuttle); and,
3) Each vehicle consumes a significant amount of space that is not easily used for any other
purpose while the vehicle is present.
7
PATH Research Report: Enhanced Transit Strategies
Autonomous garages provide the optimum convenience to users while making the most efficient
use of land area ( see www. roboticparking. com). Unfortunately, the cost of autonomous garage
implementation is typically prohibitively expensive.
Other smart parking options include intelligent parking space management, automated fee
payment, and driver information services. These ITS technologies aid to improve the traditional
method of utilizing a parking structure. The integration characteristics of various smart parking
options will be discussed in detail in later sections.
2.2.3. Low- Speed Transportation Modes
In recent years, a number of low- speed transportation modes have become very popular for a
variety of applications. These include small Neighborhood Electric Vehicles ( NEVs), electric
bicycles, scooters, CyberCars ( see www. cybercars. org), and the Segway Human Transporter
( HT). All of these modes can provide a high degree of mobility in constrained areas ( e. g.,
university campuses).
While transportation in the U. S. is dominated by automobiles, there are numerous other
transportation modes and products that are relatively new in the marketplace. While some of the
technology and vehicles have been available for many years, regulations, policies and
manufacturing characteristics have made these mobility options new to the TOD market.
NEVs are an example of vehicles that have a long history in the golf industry, but they are
relatively new to the transportation market as a mobility option ( see examples of NEVs in Figure
2.1). Recent legislation in California has allowed low- speed vehicles to travel on roads within
California as long as the posted speed limit is at or below 35 mph ( many other states have similar
legislation).
Figure 2.1. a) A four- seat neighborhood electric vehicle; b) a two- seat NEV; and c) a utility NEV.
Electric bicycles and/ or electric scooters are certainly not a new technology, but recent
advancements in technology allow them to be considered as an innovative mobility mode that
can be made part of a TOD development. These transportation modes have received limited
attention in previous years due to their relatively low performance. In recent years, their
improved power/ weight ratios have allowed for increased performance and market acceptance.
Figure 2.2 ( below) displays several electric bike/ scooter options currently being marketed. The
electric bikes are power assisted at speeds up to 18 mph and last approximately one hour. The
three- wheeled scooter offered by ZAP has a top speed of 12.5 and a maximum range of 15 miles.
The folding two- wheeled scooter has a maximum speed of 13 mph and maximum range of 8
miles ( www. zapworld. com).
8
PATH Research Report: Enhanced Transit Strategies
Figure 2.2. a) electric bicycle; b) electric 3- wheel scooter; and c) electric 2- wheel scooter.
One of the most innovative and interesting new mobility modes is the Segway Human
Transporter ( HT). The gyroscopic- balanced, two- wheeled electric scooter platform allows for
quiet and efficient personal mobility in pedestrian- oriented areas. The Segway HT mode is the
closest power- assisted mobility option to walking that is currently available. The footprint of the
Segway HT is approximately two square feet. This is nearly equivalent to the amount of space
occupied by a standing individual. Figure 2.3 ( below) shows the Segway HT with a maximum
speed of 12.5 mph and a maximum range of 24 miles ( www. segway. com).
Figure 2.3. Segway HT operating on a city sidewalk.
9
PATH Research Report: Enhanced Transit Strategies
The use of these “ new” transportation modes on sidewalks, pathways, and public areas has been
an issue of much discussion. The Segway HT has received significant attention and is the topic
of many city and state regulations [ Rodier et al, 2004; Shaheen, 2003].
It is important to note that the types of vehicles used can play a significant role in marketing a
transit- oriented development. If the vehicles are unique, new, and fun- to- drive, this can be used
as a valuable marketing tool. When the vehicles are incorporated within a TOD, there are also
potential increases to transit use.
10
PATH Research Report: Enhanced Transit Strategies
3. ITS Technology and Deployment
Prior to describing the intelligent bus priority lane concept and analysis, as well as the TOD
architecture analysis, a variety of ITS technology has been investigated ( along with several
deployments) with a focus on its applicability to enhancing transit- oriented development. Much
of the technology investigation was directed on specific data communications and constraints.
Results from this technology investigation are described below.
3.1. NATIONAL STANDARDS AND GUIDELINES
As interest grew in intelligent transportation system technology in the late 1980s and 1990s,
national standards and guidelines were developed for current and future ITS applications [ US
DOT, 2005]. An overall ITS “ architecture” was defined and has been incrementally revised
throughout the years. The ITS architecture ( defined by the U. S. Department of Transportation)
categorizes and groups the wide variety of technology and their applications. The following user
services are of particular interest for TOD development:
• Travel and Traffic Management:
􀂃 Pre- trip Travel Information,
􀂃 En- route Driver Information,
􀂃 Route Guidance,
􀂃 Ride Matching and Reservation, and
􀂃 Traveler Services Information.
• Public Transportation Management:
􀂃 En- route Transit Information,
􀂃 Public Transportation Management, and
􀂃 Personalized Public Transit.
• Electronic Payments.
The Travel and Traffic Management user service bundle focuses on user services that convey
vehicle and/ or travel information to end user locations. These services typically employ
technology that gathers information associated with the transportation facility( s), transfers the
information to a suitable end point, and displays the information to the end user. The information
may be relative to a specific vehicle ( route guidance), facility ( en- route driver information), or a
region ( traveler services information). The end user may be a transportation user or a
transportation provider/ manager ( e. g., traffic operations center).
11
PATH Research Report: Enhanced Transit Strategies
The user service bundle Public Transportation Management targets ITS solutions in the transit
arena. Similar to the previously discussed services, these user service bundles primarily promote
the transfer of transit operational data to the user. The user may be the transit rider, the transit
operator, or transportation management body.
Many transportation systems and transportation alternatives are provided at a cost to the user of
the facility, vehicle, or system. Collecting a fare from the end user is often a time consuming
task. Traditional bus systems are often slowed by the fare collection procedures involved with
individuals entering the bus. In an effort to improve the efficiency of fare collection, electronic
payments have been implemented in many transportation applications. These applications range
from electronic toll booths ( electronic toll collection) to smart cards for electronic fare collection
in transit systems.
The technology bundles associated with these user services possess many similarities. The need
for database management, data transfer, and real- time data access exists for nearly every
potential application of ITS within a TOD. The flexibility, manageability, and increasing
portability of Internet- connected devices have made the Internet the primary means of data
sharing for the majority of these applications. Web- based applications are increasingly being
used to provide transit users and operators with transit system information.
3.2. SPECIFIC TECHNOLOGY
3.2.1. Wireless Communications
Critical to many ITS applications is the ability to communicate between different devices and/ or
users. A high degree of development in the mobile wireless communication arena has occurred
in recent years with the proliferation of cellular phones, personal digital assistants ( PDAs), and
other mobile computing platforms. Much of this development has been associated with the
information needs of consumers, such as messaging, sending and receiving emails, and
downloading information from the Internet. There has also been a good deal of activity in the
communications arena of ITS. Five general types of communications linkages have been defined
for ITS, which include:
􀂃 Wide Area Broadcast Communications,
􀂃 Wide Area Two- Way Wireless Communications ( e. g., cellular),
􀂃 Dedicated Short Range Communications,
􀂃 Vehicle- to- Vehicle communications; and
􀂃 Wireline communications [ US DOT, 2005].
These communication linkages are being developed for a variety ITS applications for a range of
purposes, such as safety, remote diagnostics, maintenance, and entertainment. In general, ITS
applications have different communication requirements in terms of bandwidth, latency, and
quality of service ( QoS). For example, vehicle- to- vehicle communications in an automated
highway system scenario will require local high bandwidth communications, while applications
such as remote emergency diagnostics will need a low- bandwidth, highly available connection. It
12
PATH Research Report: Enhanced Transit Strategies
is important to note that the wireless network architecture developed for personal data
communication needs ( e. g., internet- capable mobile phones) will not necessarily be able to
satisfy all ITS communication requirements. As a result, specific wireless communication
architectures and methods are being developed and tailored for various ITS applications ( e. g.,
see [ Bana & Varaiya, 2002], [ Lee et al., 2001], [ Punnoose et al., 2001], and [ Munaka, 2001]).
Wireless communications will play a significant role in transit- oriented developments,
particularly in communicating information between users, the system, and vehicles. Much of the
communications needs make use of the Internet, since it is often widely available and a variety of
communication protocols have already been established. A variety of architectures are applicable
for TODs, using the Internet as the backbone for communications. For example, an architecture
for generic local communications between a “ system” and vehicles is shown in Figure 3.1
( below). This architecture is useful for vehicle ( or any other shared resource) access control, as
well as for uploading and downloading vehicle information. This architecture is not well suited
for real- time applications unless the resources ( vehicles in this case) do not travel far from a
local short- range communications unit.
Internet
Users: web- based access over Internet
System
Management
Server
local short- range
communications unit
local short- range
communications unit
Figure 3.1. Generic local communication architecture.
Another example communications architecture is shown in Figure 3.2 ( below). In this figure, a
generic, wide- area communication architecture is illustrated. In this case, resources ( e. g.,
vehicles) are not required to be at a designated location to communicate with the system. Instead,
cellular based communications can be used to send messages between the system and the
resources. Cellular Digital Packet Data ( CDPD) and General Packet Radio Service ( GPRS)
communications, considered as wireless Internet protocol ( IP) networks, are now widely
accepted standards in North America. They primarily provide packet data service for mobile
users by automatically using idle cellular phone channels to send packet data traffic. As such,
CDPD and GPRS have been the primary target of ITS applications that require wide- area data
communications. A mobile- end system communicates with the CDPD or GPRS network via a
19.2 kilobits per second or greater raw duplex wireless link, which is shared by several mobile
end systems. Packets from network to end systems are broadcast, thus establishing a
connectionless downlink. For the reverse direction or uplink, CDPD follows a traditional slotted,
13
PATH Research Report: Enhanced Transit Strategies
non- persistent Digital Sense Multiple Access protocol ( DSMA/ CA). Additional intelligent
wireless techniques, such as frequency hopping, radio service ( RS) code, roaming, and dynamic
channel relocation are used to provide a fairly robust data channel [ Lin, 1997]. When
implementing such a wide- area communication architecture, a monthly subscription fee must be
paid. Further, a wide- area cellular system will always have a certain degree of data packet loss
and data packet latency, which might affect shared- use vehicle system operations ( see [ Barth et
al., 2002]).
Hybrid communication architectures are also possible, as shown in Figure 3.3. This type of
architecture is particularly well suited for the multi- nodal systems where short- range
communications is used for resource access control, and wide- area communications is used for
relaying resource status information. Data packet loss and latency issues become less important
in this architecture since there is redundant communications at the different nodes.
Internet
Users: web- based access over Internet
System
Management
Server
wide- area
wireless
network
( e. g., CDPD)
Figure 3.2. Generic wide- area communication architecture.
Internet
Users: web- based access over Internet
System
Management
Server
local short- range
communications unit
local short- range
communications unit
wide- area
wireless
network
( e. g., CDPD)
Figure 3.3. Generic hybrid communication architecture.
14
PATH Research Report: Enhanced Transit Strategies
There can be many variations of the generic communication architecture examples given above.
In general, the pros and cons of these architectures are given in Table 3.1 ( below).
Communication Architecture Advantages Disadvantages
Local, Dedicated Short- Range Communications
( Figure 3.1)
• Low cost
• Low data packet loss
• Low latency
• High bandwidth
• Resources ( vehicles) can
only communicate at
stations
• Automated Vehicle
Location ( AVL) and
system messaging are not
possible
Wide Area, Cellular Communications ( Figure
3.2)
• Communications over
large areas
• AVL and system
messaging are possible
• Monthly subscription fee
required
• Non- trivial data packet
loss
• Non- trivial data latency
• Low bandwidth
Hybrid Communication Architecture ( Figure 3.3) • Communications over
large areas
• AVL and system
messaging are possible
• Redundant
communications at
stations
• Monthly subscription fee
required.
Table 3.1. Advantages and Disadvantages of Shared- Use Vehicle System Communication Architectures
3.2.2. Resource Access Control
When applying intelligent transportation system technology to shared resources ( e. g., vehicles,
lockers, seats, etc.), much can be gained by equipping the resources with on- board electronics,
especially if they are mobile. There are four primary functions that on- board electronics can
provide, namely: 1) resource access control, 2) resource data acquisition, 3) automated location
capability if the resource is mobile, and 4) on- board navigation and user/ system messaging. In
general, each of these functions can be integrated into a single black “ box” that is installed and
interfaced in the resource.
It is important to note that in many cases, installing these type of electronics may be “ overkill”
for the resource at hand. For example, it does not make sense to install elaborate electronic
devices that are more costly than the resource itself. More detail on this issue is provided in
Section 5.
That being said, having some type of resource access control improves user convenience and
system security ( potentially leading to lower insurance premiums). Minimum hardware elements
that are required for smartcard- based resource access control include a card reader ( e. g., applied
wireless identification ( or AWID)) system, which is used by several of the largest U. S.
carsharing organizations) and an interface to the vehicle’s door lock circuitry. When a user
waves his/ her smartcard by the reader, and the card is recognized as valid, the user is granted
access. That simple functionality can be implemented with discrete hardware components, not
requiring any processor. However, if a smartcard- exclusive- access methodology is used, then the
15
PATH Research Report: Enhanced Transit Strategies
sophistication of the hardware increases. In this case, user codes must be transmitted between the
system and resources so that only valid users can access the resource at the proper times. With
that added level of sophistication, typically a microcontroller or microprocessor is required to
store code variables and carry out preprogrammed state machines to implement proper
sequencing. Adding a keypad system for PIN entry does not significantly complicate the
microcontroller system, other than adding an additional hardware component to the overall on-board
electronics.
Coupled with reservations and/ or on- demand check- out procedures, there are several different
ways to control resource access:
Lockbox: All users of the resource can carry a single key that allows access to a lockbox located
at the resource location. In the lockbox, the keys of the different resources are available. Many
systems have taken this a step further by using common smartcards to access the lockboxes.
Common Key: In this scenario, all of the shared resources are keyed so that a single key can be
used for all resources. All users then have a copy of the same key and can access any of the
resources.
Smartcard Open Access to All Resources: Instead of a common key, on- board electronics ( i. e.,
card reader secured to a lock mechanism) can be used to read smartcards issued to the users. In
this scenario, all resources would unlock using any system smartcard. This method, along with
the common key and lockbox methods, depends on users following an honor system to enforce
reservations, since any user can access the resource at any time.
Smartcard Exclusive Access for Specific Users: Similar to above, smartcards are issued to
users. Each smartcard has a specific code, and when resource access is requested, only the
designated smartcard ( with the associated PIN code) can release the requested resource for use.
This resource access control requires that the smartcard code be transmitted to the resource prior
to the time of access for that user.
Smartcard Exclusive Access for Specific User with PIN Confirmation: This method is similar
to the above, where smartcard codes are used to enable specific user access. However, an
additional step is required in that once the user is at the resource, he/ she has to enter a personal
identification number ( PIN) on an input device ( e. g., message display terminal) to enable the
resource. This is similar to bank automated teller machines to help prevent fraudulent use of lost
or stolen cards.
In all of the smartcard options, key “ fobs” ( i. e., small devices that can hang from a key chain)
can also be used. Furthermore, PDAs or other wireless devices could be used for keyless access
by performing short- range communication ( e. g., infrared) with the resource.
All of these resource access solutions have tradeoffs in convenience, security, and cost. Figure
3.4 ( below) illustrates qualitatively how each access method compares in terms of security and
cost. The lockbox technique provides a small amount of security in that users have to go through
an extra step to gain access to the resource keys. The common key method is the least secure
method, since any lost key could be found and used. The smartcard- open- access method
provides a small increase in security since a person who finds a lost card would not necessarily
16
PATH Research Report: Enhanced Transit Strategies
know how to use it. The smartcard- exclusive- access method provides significantly more security
but at the cost of requiring the ability to communicate smartcard codes to the resource. The
smartcard- exclusive- access- with- PIN provides the most security and has the added cost of
requiring a PIN input device.
lockbox
security
cost
lloockbboox
common
key
smartcard
open
access
smartcard
exclusive
access
smartcard
exclusive
access & PIN
Figure 3.4. Cost and security comparison for various ITS resource access technologies.
Figure 3.5 ( below) illustrates the tradeoff between user convenience and cost. The lockbox
method detracts from user convenience in that participants must perform the step of accessing a
lockbox that may be inconveniently located. The common key method is very convenient for the
user, but there is some cost involved in having all resources keyed the same. The smartcard-open-
access and exclusive- access are equally convenient to the user. The smartcard exclusive
access- with- PIN requires an extra step prior to gaining full access to the resource and is therefore
somewhat less convenient.
user
convenience
cost
lockbox
common
key
smartcard
open
access
smartcard
exclusive
access
smartcard
exclusive
access & PIN
Figure 3.5. Cost and convenience comparison for various resource access technologies.
17
PATH Research Report: Enhanced Transit Strategies
3.2.3. Trip and/ or Resource Performance Data Acquisition
Another important function that on- board vehicle electronics can provide on any transportation
mode is the ability to automatically record trip data. These data can then be used at a minimum
for billing purposes and resource allocation analysis. In many low- technology solutions ( e. g.,
such as those used for shared- use vehicle systems), users are typically asked to complete a trip
log or diary, recording the time when the resource was checked- out and checked- in along with
the trip mileage ( if applicable). Collecting and entering these data can be time consuming for
operations. Further, this system also relies on a customer honor system. On- board electronics can
be programmed to automatically record the same parameters by interfacing with the resource
( e. g., if it is a vehicle, then we can detect usage) and using an on- board real- time clock. These
data can simply be stored and downloaded at a later time by system management personnel ( e. g.,
once every several weeks).
Alternatively, this resource usage information can be transmitted back to the system using
wireless communications. If electronics are attached to a shared resource for gathering a minimal
set of use parameters ( i. e., in a car: trip duration and trip distance), it is relatively straightforward
to extend this data set to include other useful pieces of information. Additional parameters may
include energy use, and for a vehicle, door open/ close signals, gear selection, etc. Another
valuable data parameter for mobile resources is location information, described below. It should
be noted that in the early stages of any system deployment, it is often desirable to collect a wide
range of data to document net system benefits.
3.2.4. Navigation Systems and Automated Vehicle Location Capability
In the last decade there has been a significant amount of progress in developing in- vehicle
navigation systems that help drivers efficiently reach their intended destinations. These systems
rely on electronic maps in conjunction with sensor systems ( e. g., Global Positioning System
( GPS) receivers) and associated navigation algorithms. In- vehicle navigation systems began as a
novelty offered only in rental cars and high- end luxury vehicles. However, the technology has
improved and associated costs have fallen, resulting in a wide range of vehicle navigation
systems that can be purchased separately or as part of an option package, which are increasingly
available to new car buyers. Similar progress has been made in the transit and fleet management
arena, where many Automated Vehicle Location ( AVL) systems are employed to track and
manage fleets such as buses, taxis, and delivery vehicles.
In general, vehicle navigation and AVL tasks can be broken into three scales: 1) macroscale, 2)
microscale, and 3) mesoscale.
Macroscale— the macroscale level generally considers a large roadway network as
consisting of links ( roadways) and nodes ( e. g., intersections). Specific link and node
attributes define how the network is connected together and what the general features are
of the different links/ nodes ( e. g., position, length, number of lanes, capacity, speed limit,
etc.). Macroscale navigation usually consists of finding a particular path between two
nodes in the network. This path is usually based on some optimality, such as shortest
distance or shortest traverse time. Dijkstra’s algorithm [ Chabini and Lan, 2002] is a
prime example of a solution to the macroscale route- planning problem.
18
PATH Research Report: Enhanced Transit Strategies
Microscale— the microscale level typically considers navigation at the vehicle level and
is concerned with tasks such as lane- keeping, as well as detecting and avoiding obstacles.
At this level, there is no consideration of the ultimate or intermediate goal on the route.
The driver generally carries out these tasks; however, there has been a significant amount
of research in automating many of the navigational tasks at this level, such as the work
performed for automated highway systems ( see, e. g., [ Connolly & Hedrick, 1999;
Hatipoglu et al., 2003; and Lu & Tomizuka, 2002]).
Mesoscale— the mesoscale level is a level in- between the micro- and macro- scales and
considers vehicle operation at the link- level. A particular link may have a variety of
features: multiple lanes, turn pockets, off- ramps, etc. From a navigational point of view,
mesoscale route planning is generally concerned with vehicle maneuvers, such as
passing, pulling off to the side of the roadway, moving out of the way of emergency
vehicles, merging in and out of specialty lanes ( e. g., high occupancy toll lanes), and
choosing the correct lane to exit. A link- based planning algorithm may be concerned with
when, where, and how lane changes are made with respect to a planned course change
( e. g. turn, freeway exit) or the current traffic situation.
Most of the navigational and AVL research to date has been at the macroscale and the
microscale. For in- vehicle navigation and AVL at the macroscale, sensors with positional
accuracy of approximately 10- 20 meters are sufficient. For automated microscale operations,
higher resolution sensors and actuators currently exist, but they are costly and have only been
proven in controlled environments ( e. g., automated highway systems, see [ Hedrick et al., 1994;
Horowitz & Varaiya, 2000]).
To date, very little research has been performed on mesoscale navigation tasks. Two of the
primary reasons for this are:
1) Only recently has low- cost sensor technology become available with positional accuracy
to 1- 3 meters ( e. g., differential GPS ( DGPS) receivers); and
2) Today’s digital road network data have sufficient accuracy and features for macroscale
navigation; however, they are insufficient for many mesoscale navigation tasks.
Now that it is possible to obtain sensors that have improved spatial resolution ( e. g., 1- 3 meters
using DGPS) at a reasonable cost for a vehicle, newer mesoscale navigation and AVL systems
are being developed.
For many transportation modes, it is very useful to have location information. For example, in
multi- nodal, shared- use vehicle systems where there are many one- way trips, having knowledge
of vehicle locations at any time as well as past trajectories is valuable for keeping the number of
vehicles balanced across multiple stations. Further, recording errand destination location
information can be valuable in determining where new stations should be placed. Location
information can be acquired using GPS receivers described above or by using other techniques,
such as land- based radio triangulation. The location and trajectory data need not necessarily be
transmitted in real time, it may be sufficient to record the data to be downloaded at a later time
( e. g., ignition on- and- off). AVL systems are often used on buses to help manage the fleet.
However, there are certainly privacy issues associated with AVL systems installed on ( semi-)
19
PATH Research Report: Enhanced Transit Strategies
private vehicles, i. e., those that are part of a shared- use system. Care must be taken to separate
private user data from vehicle location data in any type of analysis.
3.2.5. System Messaging
Additional functionality can be added to on- board electronics, such as integrating on- board
navigational aids that assist passengers with directions to their destinations or other information.
Also, it can also be beneficial to have system messaging capabilities so users can send/ receive
messages to the system for both emergency and non- emergency related reasons. This added
functionality can be beneficial for users and overall system operations.
3.2.6. System Management
The heart of many advanced- technology systems is the system management component. The
system management component performs various functions, depending on the system
architecture. Central to system management is usually a database consisting of users, resources
( e. g., vehicles), reservations, and trip information. Various functions that act on this database
include, but are not limited to: reservations management, check- out and check- in processing, trip
data logging, resource management ( and maintenance), and accounting ( i. e., billing). Not all of
these functions are required, and many of the functions may be spread out across different
computer platforms. Further, all of the functions may be tightly integrated automated processes;
while in other systems, some functions may be loosely coupled and/ or non- automated.
3.2.7. Reservation Management
In many transportation systems, the ability to make reservations is becoming increasingly easy.
In a low- technology implementation, a user can call a reservation center ( system management
center) and request a particular resource. An operator then checks previous reservations for the
resource( s) of interest, and if a time slot is available, the reservation is recorded. Over the last
several years, there has been significant development and proliferation of automated reservation
systems throughout society in general. For example, lodging, traditional car rental, and the
airline industries now employ automated reservation systems that can be accessed both from the
phone ( entering data via a touch- tone pad) and from the Internet. For transit- related services, it is
a natural fit to have both phone- and/ or Internet- based automated reservation systems. Generic
automated reservation systems can easily be modified for these systems, little specialization is
required for this implementation. Most on- line automated reservation systems show a calendar
with dates and times for which there are available vehicles and have a simple intuitive interface.
Reservations provide users with the comfort and security of knowing that their resource is
available for them at a specific time and place. Reservations are also useful for system
management, allowing the system to maximize resource use throughout the day.
Although reservations can provide user security and can enhance system operations, many
resource usage ( e. g., vehicle trips) in our lives are not planned well in advance. Often there is a
need on a walk- up, “ on- demand” basis. On- demand access to shared resources provides high
convenience to users; however, it places additional burden on system management to satisfy user
demand. Pure on- demand systems exist today ( i. e., systems operating without any reservation
capability). In pure on- demand systems, a “ check- out” process in which participants use a kiosk
20
PATH Research Report: Enhanced Transit Strategies
terminal located near the resource can replace the reservation process. As an example, Figure 3.6
( below) shows a touch- screen kiosk terminal located in a small building near shared- use electric
vehicles. The check- out process in this case usually involves going through a few input data
screens that are required for checking out a vehicle. Once the check- out request is complete, the
user can go to the appropriate vehicle, obtain access, and carry out the requested trip. In some
resource systems, a kiosk terminal may not be necessary; in this case, the user simply approaches
an available resource and performs the check- out and resource access process in one step.
For the on- demand check- out of resources, going first to a kiosk terminal may seem like an
unnecessary step in the overall process; however, there are several cases when a kiosk terminal
proves valuable. For example, if there is a set of homogeneous resources located at a single
location, then the kiosk computer, running system management algorithms, can play an
important role in the resource selection process. If all of the resources are the same and can
satisfy all needs, then other factors can be used in the resource selection process. For example, in
a shared- use vehicle system, choosing the vehicle with the most appropriate fuel level or rotating
vehicle use so that all vehicles are used approximately equally over time.
The process of going to a kiosk prior to accessing a resource can be circumvented through the
use of wireless- enabled PDAs or Internet- capable cell phones. In this case, a user would simply
access a website that performs the resource check- out process without going to a stationary kiosk
terminal.
Figure 3.6. Touchscreen kiosk terminal ( located inside small building) used to check- out shared- use vehicles
( electric pickups and electric city cars).
3.2.8. Accounting Systems
An important part of any system management is the ability to access data logs for billing
purposes. Further, it may be necessary to evaluate resource use based on a number of factors.
21
PATH Research Report: Enhanced Transit Strategies
Various queries and filters can be designed to quickly sort such data. User billing can be handled
as a standard back- office operation, which is prevalent on today’s Internet.
3.3. IMPLEMENTATIONS
A thorough evaluation of ITS technology associated with shared- use vehicle systems and smart
parking has been carried out to provide a detailed understanding of specific data communications
and constraints that need to be considered for TODs. As described previously, there are three
basic shared- use vehicle system models. They include neighborhood carsharing, station cars, and
multi- nodal shared- use vehicles. Recently, the first two models have advanced beyond their
original visions, largely due to advanced technologies ( e. g., electronic and wireless
communication systems) that facilitate system management and vehicle access. Thus, the initial
carsharing and station car concepts have evolved to include common elements of each model
( e. g., commuter carsharing).
3.3.1. Carsharing
Today’s typical carsharing organization places a network of shared- use vehicles at strategic
parking locations throughout a dense city ( see Figure 3.7). Members typically reserve shared- use
vehicles in advance. At the time of the rental, the user gains access to the vehicle, carries out her
trip, and returns the vehicle back to the same lot she originally accessed it from ( this is also
known as a “ two- way” rental because the user is required to rent and return a vehicle to the same
lot during one continuous rental period). Participants pay a usage fee ( typically based on time
and mileage) each time a vehicle is used. The carsharing organization as a whole maintains the
vehicle fleet ( including light trucks) throughout a network of locations, so users in
neighborhoods and business areas have relatively convenient vehicle access. Usually there is
also a small monthly subscription fee or a one- time deposit or both.
shared car
parking
shared car
parking
shared car
parking
Figure 3.7. Neighborhood carsharing model.
Internationally, carsharing organizations are the most prevalent type of shared- use vehicle
system. The vehicles are most often placed in residential neighborhoods; less frequently, they are
located in downtown business areas and rural locations. To summarize, the premise of carsharing
is simple: Short- term usage and vehicle costs are shared among a group of individuals. Lots are
22
PATH Research Report: Enhanced Transit Strategies
located so carsharing users can conveniently access vehicles for tripmaking. Often carsharing
results in increased transit ridership ( as well as other alternative modes, such as biking), as users
become much more conscious of the individual costs of each automobile trip.
3.3.2. Station Cars
Another shared- use vehicle system model is known as “ station cars”. A typical station car
scenario is depicted in Figure 3.8 ( below). When station cars are placed at major rail stations
along a commute corridor, they can serve as a demand- responsive transit feeder service on both
ends of a commute ( see [ Shaheen, 2001]). For example, a user can drive a station car from home
to a nearby transit terminal, parking it at or near the station while at work. The user then
commutes by rail or bus to their destination. After arriving at their destination station in the
morning for work, a second station car could be rented to travel from the station to their office,
and during the day the individual also might use that same vehicle to make business and personal
trips throughout the day. In the evening, the user again drives the station car to travel from work
to the station. At the end of the transit commute, this same individual takes another station car to
drive home. In this scenario, “ reverse” commuters often use the same dedicated station car for
their station- work/ station- home trips. Furthermore, other users could also make non- commute
trips during the day when the vehicles would otherwise sit idle at a station [ Bernard & Collins,
1998].
STATION STATION
school
home
office
school
home
office
SSTTAATTIIOONN sscchhooooll
hhoommee
ooffffiiccee
sscchhooooll
hhoommee
ooffffiiccee
SSTTAATTIIIOONN
SSTTAATTIIOONN
Figure 3.8. Station car model.
3.3.3. Other Shared- use Vehicle System Models
A more generalized shared- use vehicle system is one in which the vehicles are driven among
multiple stations or nodes to travel from one activity center to another. Such systems may be
located at resorts, recreational areas, national parks, corporate & university campuses, and
TODs. For example, a user may arrive by rail or bus, then rent a shared- use vehicle to drive from
the station to a corporate site, hotel, or residence, as depicted in Figure 3.9 ( below). Later on, the
same individual may travel from the hotel to a shopping mall or other attraction. In this way, the
trips are more likely to be one- way each time in contrast to the typical roundtrips made in a
traditional station car or neighborhood carsharing program. Users share vehicle costs and usage,
similar to carsharing.
23
PATH Research Report: Enhanced Transit Strategies
AIRPORT
HOTEL
RESORT SHOPS
EATERY
AAIIRRPPOORRTT
RREESSOORRTT SSHHOPPSS
SSHHOOPPSS
SHHOPS
Figure 3.9. Multiple- station shared- use vehicle model.
An advantage of a multi- station system is that vehicle trips can be “ one- way” versus “ two- way”
only. One- way rental introduces significant flexibility for users but management complexities,
including vehicle relocation. Advanced technologies can make multi- nodal systems much easier
to manage and cost effective as well.
The most effective configuration of a shared- use vehicle system within a TOD will be a function
of many variables. The integration of other compatible transit options can influence the overall
role of the shared- use vehicles significantly. The shared- use vehicles may be dedicated solely as
a transit feeder service. Other alternatives may include a TOD with high internal mobility via a
shared- use vehicle system, and the transit station being one of many potential destinations within
the shared- use vehicle system. The various shared- use vehicle system architectures have been
evaluated to explore the full range of implementation possibilities within a TOD.
24
PATH Research Report: Enhanced Transit Strategies
4. Intelligent Bus Priority Lane Analysis
Buses operating in mixed- traffic lanes experience delays due to interaction with other vehicles.
Traditional bus lanes reduce this delay in two key ways: they prevent vehicles from queuing in
front of the transit vehicle at signalized intersections, and they ensure that buses are not
competing for roadway space with private vehicles as they leave bus stops. Bus Lanes with
Intermittent Priority seek to provide the same delay reduction as traditional bus lanes by
temporarily removing private vehicle traffic in the transit lane.
To prevent queues at intersections from blocking the right- of- way of the bus, vehicles must be
removed from ( or prevented from entering) sections of a lane. This analysis considers both
conservative and liberal approaches. In both approaches, vehicles merge while discharging from
intersection queues in anticipation of preventing the formation of a queue in the bus lane further
downstream.
An Intelligent Bus Priority lane is best suited for bus routes with large headways on major urban
and suburban multi- lane arterial roads that experience medium traffic congestion during peak
periods. If traffic congestion is too heavy, the costs to other traffic of BLIP operation may be too
great; if congestion is too light, the benefits to bus passengers are minimal. Traditional bus lanes
are excellent at providing unimpeded right- of- way to bus transit vehicles, as the lane is rendered
unavailable to non- bus traffic. In situations where the bus headways ( times between bus arrivals)
are minimal, this side effect is justified. However, in situations where the headways are larger
( around 15 minutes), reserving a single lane for buses cannot be justified. However, the
alternative of operating transit vehicles in mixed traffic, results in slow and unreliable service.
Reserving the lane for buses can yield benefits of two types: reduced travel time and reduced
travel time variation. Travel time is reduced by the elimination of merge delays ( delay
experienced by buses merging back into mixed traffic lanes) and signal queue delays ( delay
imposed by queues at intersections). By removing factors prone to stochastic variation ( e. g.,
merge delay and signal queue delay) from those that influence the buses' travel time, roundtrip
bus travel time variability can also be reduced. These benefits are discussed in detail in the
following sections.
To better understand the BLIP concept, one can imagine a region of roadway that is reserved for
the bus. This region or zone starts at the bumper of the bus and extends a fixed distance ahead of
the bus. This zone is to be kept clear of non- bus traffic to ensure that the bus does not experience
any delay caused by interacting with private vehicles. In deployment, the zone reserved for the
bus will not travel continuously along the roadway, but instead travel discretely one road
segment at a time.
An example of the logic behind a BLIP activation could prove instructive: A bus traveling along
its route is equipped with an AVL system that transmits its trajectory information to a central
control system. This control system then projects the trajectory of the bus and determines at
which intersections the bus might be queued. To prevent this queuing, the system then tracks
back ( upstream) along the roadway to determine which ( and when) intersections would be
discharging vehicles that would be queuing in front of the bus. The system creates a signal plan
to ensure that signals at those intersections instruct drivers at the appropriate time that the right-
25
PATH Research Report: Enhanced Transit Strategies
most lane should be reserved for the bus. The control system performs this logic iteratively,
working its way downstream. As the bus communicates new trajectory information, the
signalization plan is updated with any changes.
A variety of roadside communication technologies can be employed to provide notification of
the intermittent lane's status, including in- pavement lights and changeable message signs
( overhead and roadside).
It should be mentioned here what this proposed concept is not intended to do. What is proposed
here will not eliminate any problems that are currently experienced with traditional bus lanes.
These problems, which include accommodating right turns and dealing with pedestrians blocking
right- turn movements, are not in the scope of this analysis. Other research is focusing on these
issues. It is important to consider this proposed concept as a bus lane that permits non- bus use
when possible. Direct comparisons to BRT should not be made.
The BLIP concept is complementary to transit signal priority ( TSP). In TSP implementations,
signal cycles are changed to give priority to transit vehicles. TSP reduces the delay caused to
transit vehicles caused by the red signals ( signal stop delay). A BLIP can be effective at reducing
the delay caused by the queue at an intersection ( signal queue delay). In implementations where
TSP and priority lanes can be paired, the bus will only need to stop for passenger boarding and
alighting. This will ultimately decrease the travel time on the route and increase the reliability of
the system by ensuring schedule adherence.
4.1. BASIC ANALYSIS
4.1.1. Scenario Description
The intelligent bus priority lane analysis uses a simplified scenario for evaluating the impacts of
the bus on through traffic. First, the analysis ignores turning traffic. It is noted below when non-trivial
turning traffic impacts the analysis. Second, it is assumed that all signals have the same
cycle length and same percentage of green time. Third, it is assumed that the signals are
coordinated, such that there is no offset between intersections: all signals turn green at the same
time. The scenario uses a free- flow speed of 60 km/ hr, and the intersections are spaced 100
meters apart. As such, the first vehicle leaving a green signal will be the first vehicle to queue at
a red signal five intersections ( 500 meters) downstream. This analysis also assumes that the
traffic demand is at capacity.
4.1.2. Supporting Concepts
Kinematic Wave Theory— this analysis uses concepts of the kinematic wave theory, also known
as the Lighthill- Whitham/ Richards ( LWR) theory [ Lighthill and Whitham, 1955; Richards,
1956]. This theory provides tested techniques for modeling traffic flow and queuing. The LWR
theory covers stationary traffic states, queue formation and discharge speeds, traffic response to
bottlenecks, etc.
Fundamental Diagram— one component of the LWR theory is the concept of the fundamental
diagram. This analysis assumes a triangular fundamental ( flow/ density) relationship for all lanes
combined as displayed in the diagram in Figure 4.1 ( below). The flow at any given point on the
26
PATH Research Report: Enhanced Transit Strategies
diagram will be expressed as a q with a subscript matching the label of the point on this diagram.
For example, the flow at point E will be expressed as qE. The diagram illustrates two “ curves”.
The first larger curve represents the roadway at “ full” capacity. The smaller of the curves
represents “ reduced” capacity roadway conditions: when one of the lanes has been reserved for
the bus and is therefore no longer available to private vehicles. The diagram illustrates the
following traffic states of interest:
A Uncongested free- flow
B Full roadway jam density
C Full roadway capacity
D Reduced roadway jam density
E Reduced roadway capacity
F Congested full roadway conditions with same flow as state E
G Congested reduced roadway with same speed as F.
Figure 4.1. Flow/ Density diagram. This specific diagram represents a three- lane roadway being reduced to a two-lane
roadway.
27
PATH Research Report: Enhanced Transit Strategies
4.1.3. Overview of Approaches
As discussed above, this analysis considers two approaches: once conservative and one liberal.
Both approaches restrict private access to the right lane at the onset of a green phase of an
intersection’s signal. The conservative approach imposes the restriction for a full cycle length.
The liberal approach imposes the restriction only long enough to ensure that private vehicles do
not queue in front of the bus. The conservative and liberal approaches are displayed in Figure 4.2
and Figure 4.3, respectively ( below).
Figure 4.2. Illustration of the conservative approach.
As illustrated in Figure 4.2, the conservative approach creates a “ rectangular” region of two- lane
traffic. Vehicles entering from the “ bottom” and the “ left side” of the rectangle are instructed to
merge when entering the restricted region. The restricted region is large enough to ensure that no
vehicle in the region will interact with the bus at some other point in time. As the figure
illustrates, notification of the road status ( restricted or unrestricted) can be communicated to the
drivers by the signals at the intersections, and each signal will display the restriction status for an
entire green phase.
This approach is likely to be less confusing to drivers, as signals will not change mid- phase, and
the restricted regions do not physically ( in space) abut unrestricted regions. The restricted and
unrestricted regions do abut temporally ( in time), and these transitions are modeled as part of the
analysis. The negative aspects of this approach are that it causes a much larger disturbance to
traffic and requires the merging of many vehicles that would not be interacting with the bus.
28
PATH Research Report: Enhanced Transit Strategies
Figure 4.3. Illustration of the liberal approach.
Figure 4.3 illustrates the liberal approach. This approach does not create a rectangular restricted
region, but instead it creates a “ slanted” restricted region that is roughly a parallelogram. The
sides of the region are defined by the trajectories of the first and last vehicle in the restricted
region, and the slope of these trajectories is free flow speed. Because all vehicles within and
neighboring the restricted region are traveling at free- flow speed, vehicles only enter the region
at the “ bottom”. As with the conservative approach, notification of roadway status can be
communicated to drivers at the intersection signals. However, this would require not only
fractional restriction notification, but switching a lane's status from “ unrestricted” to “ restricted”
and back again during a single green phase.
This approach only affects vehicles that would potentially queue in front of the bus, and
therefore minimizes the disturbance to traffic. However, there may be implementation
difficulties and driver confusion due to the restriction signalization lasting for less than a full
green phase. Additionally, the restricted region abuts in space the adjoining unrestricted regions
on both sides. This could cause additional driver confusion, as drivers could be tempted to
“ follow the lead” of the unrestricted vehicles ahead of them.
4.1.4. General Findings
In this section, the authors define two types of effects found by this “ first blush” analysis. First,
the startup effect is the capacity reduction due to the beginning of the bus along its route.
Second, the intersection effect is the capacity reduction that results from the subsequent merging
29
PATH Research Report: Enhanced Transit Strategies
movements along the route. The reason for the two effects is that the startup disturbance creates
platoons of lower flow in the traffic stream that travel upstream at about the same speed as the
bus. These low- flow platoons reduce the impact of subsequent merging movements. This is
illustrated in Figure 4.2, where vehicles leaving the “ top” of the restricted regions ( labeled E2)
continue upstream unrestricted. However, due to the conservative nature of this approach, these
vehicles are requested to merge again after queuing at subsequent signals. This merging causes
no capacity reduction, as the vehicles in question are at a less- than- capacity traffic state ( state
E3), which can easily fit into two lanes without queuing.
The first finding of this analysis is that the conservative approach has a significant startup effect,
and a moderate intersection effect. The startup effect is illustrated in Figure 4.4 ( below), where
the traffic disturbances caused by the bus beginning its route are readily apparent. Secondly, as
illustrated in Figure 4.3, the liberal approach has no startup effect and the same intersection
effect as the conservative approach.
Figure 4.4. The “ startup effect” of the conservative approach. The star indicates where the bus enters the roadway.
The intersection effect displayed by both approaches can be seen on both figures as a thin
“ ribbon” of queue that travels backwards along the roadway. The potential impact of this
backward- moving congestion is discussed in the next section.
30
PATH Research Report: Enhanced Transit Strategies
4.1.5. Macroscopic Analysis
As the above analysis has indicated, the BLIP implementation will create queues that travel
upstream when traffic demand is at capacity. Considering that subsequent buses can be delayed
by these queues, further analysis is necessary.
If we zoom out to a much larger scale of analysis, some insight can be shed on the problem. At a
macroscopic scale, the impacts caused by the signals can be averaged into a new fundamental
diagram: The free flow speed ( vf) would be the average speed of traffic ( ignoring the bus); the
maximum flow ( qC) would be the original road capacity multiplied by the fraction of green ( g/ c);
and the jam density would stay the same ( kj). ( This has the effect of reducing the backward wave
speed ( w), which should be expected since— Figure 4.3— disturbances are also slowed at the
signals when the signals are red.) This modified macroscopic fundamental diagram is displayed
in Figure 4.5 ( below).
The bus can now be modeled as an ordinary moving bottleneck. Moving bottlenecks create
different traffic conditions upstream and downstream of the bottleneck: the upstream traffic in a
congested state and downstream traffic freely flowing at a reduced volume. The interface
between these traffic states is the bus, which travels at an average speed of vB. On the
fundamental diagram, this speed is shown as a line from the origin with the slope vB, as well as a
parallel line connecting the upstream and downstream states. The downstream traffic state D will
be assumed to be the capacity of the road minus one lane, qC( n- 1)/ n, which is a conservative
estimate of the flow that will discharge from the bottleneck. The upstream state U is determined
by following the line of slope vB from state D to the congested branch of the diagram.
If the road was infinitely long and there was an infinite demand waiting to enter, the introduction
of a single BLIP bus would result in the beginning of the roadway being predominately in state
U. This is illustrated in Figure 4.5b. Therefore, we can consider the flow at this state ( qU ) to be
the capacity of the single- bus BLIP system on a very long street. It should be noted that the state
D would be the traffic state resulting from a dedicated bus lane implementation, and its flow qD
is significantly less than qB. Also, it should be reiterated that the traffic state U is a function of
the bus speed; therefore, increasing the average speed of the bus ( vB), increases the capacity of
this simplified single- bus system. Finally, it should be obvious that the ideal application of a
BLIP implementation is in a situation where the traffic demand is somewhere between qD and
qU.
31
PATH Research Report: Enhanced Transit Strategies
Figure 4.5. Fundamental diagram and time- space plot for the macroscopic view: infinite roadway and a single bus.
Extending the analysis to more than one bus presents complications, as subsequent buses could
be affected by queues created by previous buses. Luckily, roads and bus routes are not infinitely
long, and the complications are inconsequential. Figure 4.6 ( below) illustrates the situation
where the BLIP lane makes up a portion of length L of the roadway in question. The
fundamental diagram for this situation, Figure 4.6a, indicates the traffic demand state A. The
time/ space diagram in Figure 4.6b shows that the downstream reduced capacity state D meets
with and cancels out the congested upstream traffic state U from the previous bus. We see that
the flow of state A ( qA) can be sustained for as many headways as necessary, as long as qA < qU.
And this is true independently of L and H. Thus, we can think of qU as the car- carrying capacity
of the system.
32
PATH Research Report: Enhanced Transit Strategies
We assumed in the construction of Figure 4.6 that the headways are so short that the clearing
wave between states U and D does not reach the upstream end of the BLIP section— even if the
upstream demand is qU . Note from the figure that the lower bound of the time interval following
a bus arrival at the upstream end of the BLIP until the passage of its clearing wave is:
⎟ ⎟⎠
⎞
⎜ ⎜⎝
⎛
= +
v w
T L
B
1 1 .
Thus, the wave cannot reach the upstream boundary if H ≤ T . This is the condition for capacity
qU to be achieved. For typical systems, the factor in parentheses above relating T to L should be
on the order of 10 min/ mile. Hence, the maximum headway for a ( short) two- mile BLIP is ( long)
about 20 min. We expect most BLIP applications to satisfy this condition: H ≤ T . Fortunately, if
the condition is not satisfied, the car- carrying capacity is greater. In this case, as illustrated in
Figure 4.7 ( below), the maximum entry flow in each headway cycle alternates between qU ( for T
time units) and qA ( for H- T units). Thus, the complete capacity formula is:
BLIP car- carrying capacity approximation:
maximum flow of cars ≅ qU ( T/ H) + qA( 1- T/ H) , if T < H. ( 4.1a)
≅ qU , otherwise. ( 4.1b)
33
PATH Research Report: Enhanced Transit Strategies
Figure 4.6. Fundamental diagram and time- space plot for macroscopic analysis, showing two buses. Note the
downstream traffic state ( D) from the second bus cancels out the congested traffic state ( U) from the first.
This formula is an approximation based on our “ zoom concept”. It assumes that the section of
interest has many blocks and that a bus- headway includes many cycles. If these conditions are
violated, then the approximation is invalid. But then, one would not be considering BLIP lanes.
It should be obvious from Figure 4.6 that if the flow of state A ( qA) increases and exceeds qU, the
cancellation effect is removed, and headways greater than T would be necessary to accommodate
such flow. Figure 4.7 ( below) illustrates the situation. In this case, the congestion caused by the
34
PATH Research Report: Enhanced Transit Strategies
first bus will slow the second bus unless, as pictured, the headway ( H) of the buses is greater
than the time needed to allow the congestion to dissipate before the following bus begins its
route, indicated as time ( T) on Figure 4.7. The reader can verify that the critical headway ( for
which the sliver of state “ A” in Figure 4.7 disappears) is the value of H for which ( 1a) yields qA.
( This is an alternate way in which ( 1a) could have bee derived.)
Figure 4.7. Fundamental diagram and time- space plot showing macroscopic analysis where demand flow qA is
greater than congested flow qU.
35
PATH Research Report: Enhanced Transit Strategies
To summarize the conclusions from this macroscopic analysis, we see that a BLIP
implementation can accommodate any car flow qA less than qU – independently of the BLIP’s
length or the bus headway. But, higher car- flows are possible if the bus headways are greater
than “ T”; this critical time is roughly estimated at about 10 minutes per mile of BLIP.
4.2. DETAILED ANALYSIS
4.2.1. Analysis Overview
The following detailed analysis explores boundary conditions for feasibility of the liberal
approach. Since it was determined above that the intersection effect is the same for both
approaches, the resulting formulae will also apply to determining the feasibility of the
conservative approach. The liberal approach follows a rule that traffic merges upstream of a
potential bus interaction while discharging from a queue. Once the queue has cleared, traffic is
no longer instructed to merge. This analysis determines the queue clearance time of a signal,
which is a function of the offset between the signal and the next upstream signal. As such, the
authors first explore calculating the effective offset of a signal, and then determine the queue
clearance time. Finally, the impacts in space and time are evaluated.
Figure 4.8 ( below) displays a time- space diagram that provides an example for a three- lane
roadway. The vehicles denoted by the solid trajectories are the first and last to queue at the
intersection where the bus is expected. These vehicles and any in between will queue at the
upstream signal as normal, but they will discharge from that queue in only two lanes. This will
ensure that the vehicles at the downstream intersection queue in only two lanes. This leaves a
lane open for the bus which, represented by the broken line, can jump the queue and pull up to
the stop line.
Again, in this scenario, once the queue at the upstream intersection has dissipated, vehicles
arriving at the intersection are permitted to use all lanes. If the vehicles arriving after the queue
has dissipated are anticipated to interact with the bus, they will have already merged at an
intersection even further upstream. If not, they will either arrive at the downstream intersection
after the bus has passed or they will be stopped at an intermediate intersection.
4.2.2. Supporting Concepts
System Inputs— the following variables will be used throughout the detailed analysis.
qX Flow at traffic state X
g Green time
c Cycle length
t Time. Used to illustrate " specific" times ( t1, ti, ti+ 1, etc.)
tX Time of interest in traffic state X
O Offsets, expressed in time units
36
PATH Research Report: Enhanced Transit Strategies
L Length of roadway segment, usually the distance between intersections.
vF Free flow speed.
Figure 4.8. Example of BLIP activation. Traffic merges from 3 lanes to 2 lanes while discharging from an upstream
signal in anticipation of queuing in 2 lanes downstream. The broken line represents the trajectory of the bus, and
the solid lines represent the first and last vehicle that will queue at the intersection where a bus is expected.
Offsets— an offset is the time difference between signal cycles at subsequent intersections.
Offsets can be expressed as absolute, relative or effective. An absolute offset ( OA) is the actual
time difference between initiations of the green phases of two signals. A relative offset ( OR) is
the absolute offset adjusted by the free- flow travel time between intersections. Relative offsets
can be positive or negative and are always between - c/ 2 and c/ 2.
O R = O A − L
v f
The effective offset ( OE) is the amount of time the red signal of an intersection is exposed to
traffic from the upstream signal.
Actual and effective offsets are illustrated in Figure 4.9 ( below). The basic equation for the
effective offset is simply the absolute value difference of the relative offset:
O E basic = O R .
37
PATH Research Report: Enhanced Transit Strategies
Figure 4.9. Comparison between actual and effective offset. The actual offset is represented by ta and the effective
offset is represented by tO. Part ( a) shows a negative actual offset; part ( b) shows a positive actual offset.
The absolute value is necessary here due to the fact that the effective offset's sign does not have
an effect on the queue length: whether the vehicles arrive at the start of the red or towards the
end, the queue length does not change. All that matters is the amount of time that the red signal
is exposed to oncoming traffic from the previous signal.
Due to the cyclical nature of traffic signals, this basic formulation must be further refined to
accommodate for the situation where the signals are anti- coordinated. In other words, if the basic
effective offset is greater than the green time provided by the signal:
O E =
O E basic O E basic < g
min( g, c − O E basic ) g < O E basic
⎧
⎨
⎩
.
This expression captures the fact that if the basic effective offset is greater than the green time
provided by the signal, the effective offset will be equal to the green time of the upstream signal.
The effective offset is useful when determining the amount of queuing at an intersection given
the coordination ( or lack there of) between a signal and other upstream signals. More
specifically, the effective offset is the time during which a red signal could be exposed to
saturation flow traffic from an upstream intersection. For example, if the actual offset is equal to
the free flow travel time between intersections, the downstream signal will turn green as the first
vehicle discharging from the upstream queue reaches the intersection, resulting in an effective
offset of zero and no queuing at the intersection.
Queue Clearance Time— mentioned above, the activation of a BLIP will be activated at an
intersection for the amount of time that it takes the queue to clear at that intersection. As
displayed in Figure 4.10, this “ queue clearance time” is defined as the elapsed time between the
initiation of the green phase and the time the last queued vehicle crosses the stop line. ( It should
be noted that this last vehicle might not have been queued when the signal turned green.) The
queue clearance time is a function of the size of the queue at an intersection, and that queue size
is subsequently a function of the traffic flow from the upstream signal⎯ the offset between the
38
PATH Research Report: Enhanced Transit Strategies
signals and the queue discharge rate. This clearance time can be determined analytically. The
size of the queue by definition is the flow that is stopped at the signal.
Figure 4.10. Queue clearance time for non- isolated ( a) and isolated ( b) intersections.
For an intersection in a series, as illustrated in Figure 4.10a, the red signal is only exposed to
flow for the duration of the effective offset, tA = OE. ( Here tA represents the time the signal is
exposed to traffic state A, which is equal to the effective offset ( OE) calculated above.) As such,
the queue clearance time, tE, for a non- isolated intersection can be calculated easily using
queuing concepts. The queue size, Nq, will simply be the flow arriving at the intersection times
the effective offset, N q C E = q ⋅ O . Here, qC is the saturation flow of the discharging upstream
signal. The same will apply to the discharge of the queue, N q = q ⋅ t E E , where qE is the saturation
flow of the signal under inspection. Setting the right- hand sides of these equations equal to each
other and solving for tE results in the following equation for the queue clearance time of a non-isolated
signal:
t E = q C
q E
t A .
For an isolated intersection with an assumed constant flow less than saturation, as illustrated in
Figure 4.10b, vehicles will be interrupted not only by the red signal but also by the tail end of the
dissipating queue, resulting in vehicles queuing for a duration of ( c- g) + tE. Using the same
method used above, the following equation can be derived for the queue clearance time for an
isolated intersection:
t E = q A ( c − g)
( q E − q A )
.
If traffic turning on to the arterial is considered, a factor will need to be added to the arrival flow
quantity.
39
PATH Research Report: Enhanced Transit Strategies
4.2.3. Other Factors
In addition to the system inputs described above, there are other factors that should be given
consideration. These factors are primarily concerned with the design and location of bus stops.
Bus stop locations are named according to their relation with the intersection: far- side, near- side,
or mid- block. An arterial using near- side bus stops has the most to gain from a BLIP
implementation, as the passenger movements can be made while the bus is stopped at a signal.
Far- side and mid- block bus stops may not gain as much overlap benefit.
Bus stops can be configured as bus bays ( or turn- outs), bus bulbs, or curb- side stops. Bus routes
along arterials with bus bays will gain more benefit ( merge delay reduction) than in- lane bus
stops ( bus bulbs and curb- side stops). This is because bus routes with in- lane bus stops do not
experience merge delays.
These factors should be considered when determining the feasibility and benefits of a BLIP
implementation. Bus routes that use only bus bays and near- side bus stops have the most to gain
from a BLIP implementation. Bus routes with far- side bus bulbs, for example, have the least.
They might only benefit from a reduced signal queue delay. Each intersection and bus stop
should be considered independently with its unique characteristics.
4.2.4. Feasibility Analysis
A series of simple calculations can be performed on an intersection- by- intersection basis to
determine whether a BLIP implementation is feasible along a given roadway segment. The
criteria for feasibility include:
􀂃 Impacts constrained in time: Implementation will not create a prolonged disturbance
over time.
􀂃 Impacts constrained in space: Implementation will not cause queues that spill back
beyond a predefined distance.
4.2.4.1. Impacts in Time
The duration of the disturbance caused by reserving a lane for traffic is localized to the merge
movements of private vehicles as they vacate the lane reserved for the bus. As stated above, this
analysis recommends that these merge movements are performed as an intersection queue
discharges. It can be easily imagined that a three- lane queue discharging into only two lanes
would have some non- trivial impact on traffic flow on the roadway.
Figure 4.11 ( below) displays a time- space diagram of the situation where a base traffic flow
( state A) queues at an intersection in three lanes ( state B) and then discharges at a two- lane free
flow ( state E). This merge process creates a new traffic state ( state F): the removal of a lane at
the intersection can be seen as a stationary bottleneck, and the discharging queue results in
different states on either side of the bottleneck: uncongested downstream ( state E) and congested
upstream ( state F).
40
PATH Research Report: Enhanced Transit Strategies
Figure 4.11. Time- space diagram illustrating merge during the activation of the priority lane. For example, vehicles
traveling in state A queue in three lanes ( state B), but they merge to two lanes as they cross the stop line ( state E). A
three- lane congested state ( F) results directly upstream of the intersection. The grey lines represent vehicle
trajectories, and the dashed line represents the last vehicle in the queued state B. Once this vehicle reaches the stop
line, subsequent vehicles proceed through the intersection in the original traffic state ( C).
The duration of the disturbance caused by the activation of a BLIP is called the relaxation time.
The starting point for determining the relaxation time of the disturbance is a queuing diagram,
such as the one displayed in Figure 4.12 ( below). This relaxation time n, expressed in either
cycles, can be determined analytically through the following supply and demand metaphor. The
demand for the intersection in question is simply desired flow during the relaxation time:
q A nc
where qA is the “ base” flow or demand, n is the number of cycles that the disturbance persists,
and c is the cycle length.
The supplied capacity of the intersection is made up of three parts:
q E t E + q C ( g − t E ) + q C g( n − 1).
The first part ( qEtE) gives the flow capacity available during activation of the BLIP at the
intersection, where qE is the reduced saturation flow, and tE is the queue clearance time. The
second part gives the number of vehicles that can clear the intersection during the remainder of
the green time after the queue has cleared, where qC is the saturation flow, and g is the cycle
green time. The third part gives the number of vehicles that can depart at saturation flow qC for
the remaining n- 1 cycles.
Setting the supply equal to the demand and solving for n results in the relaxation time, given in
number of cycles.
41
PATH Research Report: Enhanced Transit Strategies
q A nc = q E t E + q C ( g − t E ) + q C g( n − 1) => n = t E ( q C − q E )
( gq C − cq A )
.
Using this equation, decision makers can set limits on the relaxation time and determine whether
a given roadway/ bus route can support a BLIP implementation. Since the saturation flow ( qC) is
known to be greater than the reduced outflow provided under bus lane activation ( qE), the
numerator of this equation will be positive. From this formulation, it can be seen that the number
of cycles will approach infinity as the denominator approaches zero. From this, we can
determine another criterion for feasibility:
gq C − cq A > 0 => q A < g
c
q C.
That is, the demanded flow must be less than the flow capacity provided by the intersection. If
they are equal, infinite queuing will occur until traffic conditions change.
Figure 4.12. Queuing diagram showing the dissipation of the disturbance caused by the BLIP activation. The
demand for the intersection is a constant, qA, represented by the solid line. Normally the intersection has a saturation
flow rate of qC. It is obvious that the intersection can support the base demand. Under BLIP activation, the outflow
of the intersection is reduced to qE, represented by the lower dashed line. This low outflow lasts until the last queued
vehicle leaves the intersection ( after tE seconds) when normal saturation flow qC resumes. During the following
cycle, the “ disturbed flow” catches up to the expected, undisturbed intersection outflow.
The delay to other vehicles can be easily evaluated using the input- output diagram displayed in
Figure 4.12. In this example, it is clear that the delayed departures catch up to the desired
departures after one cycle. From the data used to derive the above queuing diagram, one can
easily calculate delay caused by the bottleneck: the delay is the area between the two departure
curves. This delay can be calculated geometrically or through analytical methods with a
42
PATH Research Report: Enhanced Transit Strategies
spreadsheet. This delay is one of the costs that should be considered when evaluating a potential
BLIP implementation. These costs will be discussed later in the report.
It should be noted that this delay might not be newly created delay: the interaction between buses
and private vehicles often causes delay. The delay calculated here could simply be a
representation of normal interaction delay. The determination of this depends highly on
characteristics of the roadway, including the bus stop configuration. It is possible that the delay
described above could be less than that which would occur due to normal bus- vehicle
interactions.
The impacts in time of the disturbance caused by the activation of a BLIP displayed above can
help determine the feasibility of implementing this architecture on a given bus route/ roadway
segment.
4.2.4.2. Impacts in Space
Any disturbance in traffic flow not only persists in time, but it also exists in space: traffic queues
take up physical roadway space. It might be desirable to ensure that queues caused by a BLIP
implementation do not grow beyond a certain length: for example, one may wish that a queue
does not back up into the previous upstream intersection. The length of a queue created by a
BLIP’s activation can be analyzed using time- space diagrams.
The length of a queue is a function of the red time and the arrival flow rate. For isolated
intersections, this calculation is straightforward. For intersections in series, the vehicle arrivals
depend on the offset of the upstream signal. ( For example, if the signals are perfectly
coordinated, no vehicles will arrive during the red phase of the signal.)
Figure 4.13 ( below) illustrates queues growing and dissipating at isolated and networked
intersections. For isolated intersections, as shown in Figure 4.13a, vehicles arrive in stationary
traffic state A, and the speed at which the back of the queue grows is UAB. Given that vehicles
will leave the queue in a different traffic state than they arrive, traffic state E, the speed at which
the front of the queue dissipates can be represented by UBE. The location of the back of the queue
growing for time t1 can be expressed as t1UAB. The location of the front of the queue after
discharging for a time t2 can be represented by t2UBE. Since the queue is fully discharged when
the front of the queue meets the back, the maximum queue length occurs where the two meet:
L = t 1 U AB = t 2 U BE
Additionally, we know the queue begins forming when the signal turns red and begins
discharging when it turns green. Therefore, t1 = t2 + R, where R is the red time of the cycle.
Solving these equations for t2 and then for L results in the following equation for the maximum
queue length of an isolated intersection:
L = U AB U BE
U BE − U AB
R.
Since a criteria for possible implementation if a BLIP is to ensure that the queue caused by
reduced queue discharge rate does not extend beyond a certain length, L, it is desirable to
43
PATH Research Report: Enhanced Transit Strategies
determine an upper bound for the demand flow, qA. This can be derived from the above equation
by substituting the definition for the “ interface” speeds, i. e., U AB = ( q − q ) /( k − k B A B A ), and then
solving for qA. This results in the following expression for the maximum value of qA:
q AMAX
= q E ⋅ ( k A − k E )
R
L
q E + k E − k B
Figure 4.13. Graphic illustration of impacts in space for isolated and non- isolated intersections. a) The queue length
( L) at an isolated intersection is a function of the arrival and discharge traffic states ( A and C respectively). b) The
queue length ( L) at a non- isolated intersection is a function of the traffic state ( A) and the offset from the previous
signal.
In the case of intersections in series ( non- isolated), the queue length is a function of the arrival
flow rate and the offset from the previous signal, as discussed above and illustrated in Figure
4.13b. The queue will grow at the rate UAB, while the red signal is exposed to flow from an
upstream signal, the effective offset time tA = tO
L = t A U AB .
Substituting the definition for the interface speed ( as discussed above) and solving for qA will
result in the maximum flow qAmax that can arrive at the red signal without the queue spilling
beyond our pre- defined distance L
q Amax
= L( k B − k A )
t A
.
Since this formulation is for signals in a series, the qAmax may be the saturation flow from an
upstream intersection, coming to the current intersection in platoons with flow qA, but having an
average flow significantly lower than qA. If this is the case, the average flow can be given by:
q Amax
= c
g
⋅ q Amax ,
44
PATH Research Report: Enhanced Transit Strategies
where c and g are the cycle length and green time of the upstream signal.
It should be noted that, depending on the cycle offsets and the overall traffic demand on the
arterial, the flow arriving at the signal during the red phase may or may not be saturation flow. If
this analysis predicts queues that grow to unacceptable lengths, the signal offset should be
adjusted in an attempt to ensure that the signal is exposed to a flow at a level below the
saturation flow. However, if the system is at or near capacity, this may not be possible.
4.2.5. Benefit Analysis
The benefits of a BLIP implementation fall into two categories: reduced mean travel time and
reduced travel time variation. These are explored below.
4.2.5.1. Reduced Mean Travel Time
Transit vehicle travel time is usually estimated using three factors. The first is the distance
traveled divided by the free- flow speed of the bus. The second, signal delay, is time spent
waiting at traffic signals. The third, stop delay, is the time required to stop for the discharge and
boarding of passengers. Bus Lanes with Intermittent Priority can help reduce the signal delay
and stop delay components of bus travel times.
4.2.5.2. Reduced Signal Delay
Signal delay for a transit vehicle is defined as the delay experienced at signalized intersections.
This delay can be broken into two components: signal stop delay and signal queue delay. The
signal stop delay is the delay caused by the red stop signal. The signal queue delay is component
of the delay caused by the existence of other vehicles in the queue ahead of the bus. Transit
Signal Priority ( TSP) has been proposed to help reduce signal stop delay by modifying the green
time of a given cycle period to give priority treatment to the bus. This BLIP proposal attempts to
eliminate the signal queue delay portion of signal delay.
Under a BLIP implementation, the reservation of the lane allows a bus to “ jump the queue”. The
amount of delay saved by a bus as it jumps the queue at an intersection is highly variable, and
the delay depends on the traffic volume as well as the bus arrival time at the intersection in
relation to the cycle. Figure 4.14 ( below) shows examples of time savings as a function of arrival
time. If the bus arrives just as the signal turns red, as in Figure 4.14a, there is no queue- jumping
savings; there would be no queue in front of the bus and the entire signal delay is all due to the
red signal. However, a bus with a trajectory such that, if there were no queue at all it would reach
the stop line of the intersection the instant the signal turns green as in Figure 4.14c, will gain
much benefit from jumping the queue.
The fundamental diagram in Figure 4.1 applies to this analysis, and the signal queue delay of a
bus trajectory at an intersection can be calculated, given the following parameters:
45
PATH Research Report: Enhanced Transit Strategies
c cycle length
g effective green time
A initial traffic state
B traffic state of queued vehicles
C traffic state of discharging vehicles
UAB Speed of interface between A and B
UBC Speed of interface between B and C
vf Freeflow speed of bus
t0 Time the signal turns red
x0 Location of the signal
xB location of bus at t0
Figure 4.14. Delay reduction benefits as a function of bus arrival time. The thick dashed line represents a bus
trajectory that uses a BLIP, and thin dashed line represents trajectory of bus without priority treatment. a) Bus
arrives at onset of red signal and receives no benefit. b) Bus arrives near middle of red and receives some benefit
and experiences some signal delay. c) Bus arrives at end of signal and receives maximum benefit. d) Bus arrives
after signal has turned green, and receives benefit by jumping the residual queue.
46
PATH Research Report: Enhanced Transit Strategies
Figure 4.15 ( below) shows a time- space diagram representation of an isolated intersection. The
trajectory of the bus arrives at the back of the queue at tq but proceeds to the stop line. 1 The delay
that would have been ex