ISSN 1055- 1425
November 2007
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 RTA 65A0160
CALIFORNIA PATH PROGRAM
INSTITUTE OF TRANSPORTATION STUDIES
UNIVERSITY OF CALIFORNIA, BERKELEY
Lane Assist Systems for Bus Rapid Transit,
Volume II: Needs and Requirements
UCB- ITS- PRR- 2007- 22
California PATH Research Report
Wei- Bin Zhang, Steven Shladover, Douglas Cooper,
Joanne Chang, Mark Miller, Ching- Yao Chan, and
Fanping Bu
CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS
Lane Assist Systems for Bus Rapid
Transit, Volume II: Needs and
Requirements
Wei- Bin Zhang, Steven Shladover, Douglas Cooper, Joanne Chang,
Mark Miller, Ching- Yao Chan, and Fanping Bu
Prepared by:
University of California at Berkeley
PATH Program
1357 South 46th Street
Richmond, CA 94804
Prepared for:
California Department of Transportation and
U. S. Department of Transportation
Federal Transit Administration
Final Report for RTA 65A0160
ii
iii
Acknowledgments
This report was prepared in cooperation with the State of California, Business
Transportation and Housing Agency and United States Department of Transportation
Federal Transit 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 or the Federal Transit Administration. This report does not constitute a
standard, specification, or regulation.
This project has depended on the cooperation of many participants besides the report
authors. The authors would like to thank Walter Kulyk and Brian Cronin of the Federal
Transit Administration, Don Dean and Z. Sonja Sun of the California Department of
Transportation and Matt Hardy of Mitretek for their advice and support throughout the
project. They appreciate the assistance of their PATH colleagues Tunde Balvanyos and
Han- Shue Tan during the work on the project. The project has gained greatly from the
cooperation of the participants from the partner transit properties:
Lane Transit District
Mark Pangborn, General Manager
Graham Carey, BRT Project Engineer
Stefano Viggiano
Stephen Rayack
Tim Jacobson
John Gonzales
Javier Rodriguez
Terry Bean
Nancy Nielsen
Diann Sheldon
AC Transit
Rick Fernandez, General Manager
Jim Cunradi
Aaron Priven
Huaqi Yuan
Joe Kinchen
Ken Rhodes
San Diego Transit, SANDAG, and City of San Diego
Dave Schumacher
Brian Sheehan
Miriam Kirshner
Steve Celniker
Kathy Donnelly
John Duve
Deena Smith
iv
v
Kathie Wellington
Jennifer Williamson
Lost Angeles County Metropolitan Transportation Authority ( LAMTA)
Rex Gephart
David Mieger
Kevin Michel
Edward Clifford
Joe Vicente
Kildare, Greg
DiNuzzo, Alex
Roberts, John
Pachan, James
Trudeau, George
Smith, Craig
Brewer- Smith Yvonne
Hogancamp, Robert;
Whitney, Ronald
Butler, Martha
Page, Scott
McAllester, Bradford
Inge, Carol
Goldman, Roderick
Caltrans District 11 ( San Diego)
Lynn Barton
Chris Schmidt
vi
vii
Abstract
This report defines the transit service needs that can be met by use of lane assistance
systems and the requirements that these systems must meet in order to be useful and safe.
The work is based on review of the existing literature and research on the subject of lane
assistance, combined with case studies of several transit properties that could potentially
benefit from use of lane assist systems. The project team has conducted workshops
involving participation by a broad mix of people representing the transit properties in
order to learn about the needs that they perceive, as well as both benefits and risks that
they perceive associated with use of lane assist technology. Because these transit
properties are significantly different from each other, their needs are also diverse.
Keywords: Vehicle Highway Automation, Lane assist, electronic guidance, Bus Rapid
Transit, precision docking, automatic steering
viii
ix
Executive Summary
This report defines the transit service needs that can be met by use of electronic guidance
systems and the requirements that these systems must meet in order to be useful and safe.
Guidance system functions include automatic steering of buses driving between stations
( lane assist) and automatic precision docking of buses at stations. These functions can be
implemented using a variety of sensing and reference technologies, but those
technologies are not the primary focus of the work here. Rather, this report addresses
issues that are essentially independent of the chosen technology.
The work is based on review of the existing literature and research on electronic
guidance, combined with case studies of several transit properties that could potentially
benefit from use of electronic guidance systems ( Lane Transit District of Eugene, OR,
AC Transit of the Oakland/ East Bay region of northern California, Los Angeles County
Metropolitan Transit Agency and San Diego Association of Governments). The project
team has conducted workshops involving participation by a broad mix of people
representing these transit properties in order to learn about the needs that they perceive,
as well as both benefits and risks that they perceive associated with use of electronic
guidance. The work has been based on the concept that it is necessary to understand the
perspectives of not only the transit operating agencies, but also of the bus drivers and the
passengers, since it will be important to have acceptance and support from all of them.
The report begins with a basic introduction to electronic guidance systems, including
project background, a summary of the candidate technological implementations, and the
relationship of the guidance system to the typical work of the bus driver. This raises a
variety of important issues that need to be considered in the design of any guidance
system:
- maintaining an appropriate level of driver workload
- reducing driver stress, particularly by ensuring high system reliability
- clarifying roles of the driver relative to the system
- carefully designing control transitions between driver and system.
Potential application environments for electronic guidance are described as new median
bus lanes, restriped bus lanes, narrow bridges, tunnels or toll booths, and new dedicated
busways. The technologies that could be used to provide guidance information are
classified as mechanical ( curb contact), computer vision, permanent magnets, current-carrying
wires, and GPS satellite navigation systems. Their general advantages and
disadvantages are described qualitatively.
Section 2 collects the inputs that were derived from site- specific case studies for four
transit properties at different stages of maturity in their consideration of electronic
guidance for their BRT systems, including workshops conducted at all four properties.
Lane Transit District ( LTD) is the most mature, having already seriously considered
adding guidance capabilities to the BRT system that they are introducing to the public in
late 2006. AC Transit is not as far along in the process, proceeding with the
environmental documentation for their new BRT system assuming that it will not include
x
guidance capabilities, but willing to consider adding it if it appears to be sufficiently
advantageous. Although Los Angeles County MTA already has an extensive BRT
system, they have not been seriously considering guidance applications until now.
SANDAG is still in the early stages of thinking about their BRT service and has not yet
actively considered guidance capabilities. Section 2 is organized by the issues that cut
across all four properties: specific applications, qualitative benefits and costs, system
requirements, design features and institutional challenges to deployment.
The LTD case study highlighted many of the important issues for electronic guidance
systems from the perspective of an operator that has already been thinking seriously
about the issues. They are primarily interested in electronic guidance for facilitating
narrower track and right of way on the exclusive busway parts of their network and for
precision docking to facilitate disabled access. Their concerns about electronic guidance
revolved around issues of liability, driver engagement, control transitions, maintenance of
vehicle systems ( ensuring service intervals comparable to other bus subsystems), and
maintenance of infrastructure. Other important issues specific to driver interfaces
involved the need for training to address both normal and abnormal conditions, a DVI
that provides useful feedback about the status of the system, and the constraints imposed
by a state law requiring that the driver’s hands always be on the steering wheel.
AC Transit was interested in several different applications of guidance capabilities –
precision docking on their existing Rapid line, lane assist on a new BRT service that will
operate in the roadway median, automation of bus maintenance facility and yard
operations, and maintaining speed while driving through a toll booth with narrow lateral
clearance. Their interest in precision docking is motivated by a high percentage of
disabled riders and the potential that enhanced accessibility of their mainline bus services
could reduce the need to provide costly para- transit services to disabled riders. With
precision docking, they would hope to be able to save enough time at bus stations that
they could eliminate the cost of running an additional bus or buses while providing the
same level of service. The lane assist function could enable them to save width of their
new BRT busway, thereby reducing costs and neighborhood concerns about loss of on-street
parking and could also provide a more rail- like smoothness of ride. Bundling this
with other advanced technology features such as forward collision warning and adaptive
cruise control could lead to smoother accelerations and decelerations, thereby reducing
passenger falls. Their concerns involve practical considerations of system durability
when operated over the long term on rough pavement surfaces.
Los Angeles County MTA would be interested in seeing a large- scale demonstration of
electronic guidance capabilities on a closed test track before committing to deployment
on their buses. Their priority considerations are system safety, reliability and consistency
of performance. The potential benefits they could envision are focused primarily on time
savings through station dwell time reductions and higher speeds in narrow lanes. They
have a variety of concerns that must be addressed, including the need for liability
indemnification from the system supplier, the potential for drivers to demand higher pay
based on any additional skills or training required to operate the system, worries about
the engineering work needed to design and implement the docking profile for each bus
xi
station, and worries about the ability of the system to accommodate road surface
imperfections beyond their control ( poorly maintained and severely crowned surfaces).
The key issues were addressed in a joint workshop held with representatives of the
participating transit properties to discuss and compare their requirements and needs. This
workshop brought forward the difficulty of defining a single uniform set of requirements
for all lane assist systems, since each property expressed different needs and priorities.
This is one of the primary challenges to the development of these systems. The
categories of issues that were covered included system performance ( lane tracking and
docking accuracy, ride quality), infrastructure changes needed, reliability and robustness,
maintenance of vehicle and infrastructure elements of the system, availability, driver
acceptance ( primarily fault handling), safety, public and passenger acceptance, and costs.
The most compelling benefit argument for the lane assist functions was the ability to save
enough time on the bus route that a bus could be eliminated from operations while
maintaining the same service ( operating headway).
Section 3 includes more detailed quantitative analyses of system benefits and costs. The
potential benefits of electronic guidance systems are introduced in qualitative terms, in
the categories of:
- providing rail- like image and service quality
- reducing driver stress
- reducing infrastructure construction and right of way cost by narrowing lanes
- saving stop time at bus stations
- providing enhanced accessibility to elderly and disabled passengers
- improving safety by reducing driver errors
- reducing paved surface area for exclusive busways.
Preliminary economic evaluations of electronic guidance systems in Section 3.2 focus on
the time savings from precision docking and the lane width savings from automatic
steering. These analyses are sensitive to the costs for equipping each bus, which were
assumed to be $ 100 K for small numbers of buses in the very near term, $ 14 K per bus
when produced in quantities of hundreds of buses per year, and $ 2.7 K per bus when
produced in quantities of thousands of vehicles per year in the future. The docking
analyses sought to identify the break- even points in station dwell time reduction where
the costs of the system would be matched by the savings in bus operating costs. These
ranged from 4.2 seconds per stop for Lane County to 12.3 seconds per stop for Los
Angeles ( where the study considered only a short section of three stops on their Metro
Rapid service, but would still have required equipping all the buses operating that
service). The automatic steering analyses sought to identify the break- even busway
construction costs where the costs of the lane assist system would be matched by the
reduction in surface area needed for narrower lanes. These ranged from $ 15.35 per
square foot for Lane County to $ 38.67 per square foot for Los Angeles.
The potential safety benefits of electronic guidance systems are addressed in Section 3.3,
based on a review of existing actuarial data on bus safety problems and consideration of
how the scenarios and maneuvers associated with those problems ( collisions and other
xii
incidents leading to insurance claims) would change with use of guidance systems. The
safety improvements are expected to be associated with reducing passenger boarding and
alighting problems, reducing passenger falls, and improving drivers’ focus on hazards in
the driving environment so that they can avoid some collisions.
Section 4 provides the system requirements, specifications, measures of effectiveness and
a preliminary example system safety analysis. This begins with a functional
decomposition and generic system architecture. The architecture begins at the logical
level, then proceeds to a physical architecture, independent of technology.
The categories of relevant functional requirements are identified as:
- safety
- performance ( ride comfort, accuracy, ease of operation)
- reliability
- availability
- maintenance ease and cost
- compatibility with existing infrastructure.
These are closely related to the measures of effectiveness needed to evaluate different
guidance system design alternatives:
- safety
- operational improvements ( time savings, ride quality)
- reliability
- availability
- capital costs
- operating and maintenance costs and other needs imposed on the agency
- weather limitations
- infrastructure impacts
- public perception
- driver acceptance.
Section 4.3 of the report introduces draft specifications and requirements for electronic
guidance systems. These begin at the level of system performance, addressing issues
such as accuracy of lane keeping and position indications to the driver, driver display
( DVI) contents, acceptable weather conditions for operation, ride quality, and control
transitions. Subsystem requirements include issues such as position sensing capabilities
( coverage, accuracy, resolution, update rate, environmental robustness), actuator
performance, and DVI characteristics. Infrastructure requirements and driver training
and maintenance requirements are also addressed.
The safety analyses begin with a preliminary hazard analysis, then a determination of
safety integrity level, based on the probability and severity of the identified hazards.
These hazards include environmental factors, driver actions, passenger actions,
component failures and design errors. Finally, a failure mode effects and criticality
analysis ( FMECA) is presented, based on the specified functions and individual failure
modes for each function
xiii
Table of Contents
ACKNOWLEDGEMENTS............................................................................................................... ........ iii
EXECUTIVE SUMMARY........................................................................................................................ ix
1.0 INTRODUCTION................................................................................................................... .............. 1
1.1 PROJECT BACKGROUND..................................................................................................................... .. 2
1.2 TECHNOLOGY OPTIONS ........................................................................................................................ 2
1.2.1 Mechanical Guidance .................................................................................................................. 3
1.2.2 Electronic Guidance .................................................................................................................... 4
1.2.2.1 Vision- based Guidance ............................................................................................................................... . 5
1.2.2.2 Magnetic Guidance ............................................................................................................................... ....... 6
1.2.2.3 Wire Guidance ............................................................................................................................... .............. 8
1.2.2.4 Global Positioning System ( GPS) Guidance ................................................................................................. 8
1.2.3 Comparison of different technologies ........................................................................................ 10
1.3 HUMAN FACTORS CONSIDERATIONS .................................................................................................. 13
1.3.1 Bus Operation is a Stressful Vocation....................................................................................... 13
1.3.2 Previous efforts to make urban bus operation less stressful ..................................................... 14
1.3.3 Previous Lane Assist Human Factors Research ....................................................................... 14
1.3.4 Driver Vehicle Interface ( DVI) Design Issues ........................................................................... 16
1.4 TERMINOLOGY DEFINITIONS .............................................................................................................. 17
1.5 REPORT ORGANIZATION ..................................................................................................................... 18
2.0 CASE STUDIES ............................................................................................................................... ... 19
2.1 POTENTIAL APPLICATIONS ................................................................................................................. 20
2.1.1 LTD Case Study Example .......................................................................................................... 20
2.1.2 AC Transit Case Study Example ................................................................................................ 23
2.1.3 SANDAG’s Considerations in Lane Assist and Precision Docking Systems.............................. 24
2.1.4 LACMTA’s Considerations in Lane Assist and Precision Docking Systems.............................. 25
2.2 BENEFITS AND COSTS ......................................................................................................................... 26
2.2.1 Benefits....................................................................................................................... ............... 26
2.2.2 Costs.......................................................................................................................... ................ 29
2.3 REQUIREMENTS ............................................................................................................................... .. 30
2.3.1 Performance.................................................................................................................... .......... 30
2.3.1.1 Lane- tracking accuracy .............................................................................................................................. 31
2.3.1.2 Precision Docking Accuracy...................................................................................................................... 32
2.3.1.3 Lane tracking ride quality ........................................................................................................................... 33
2.3.2 Safety......................................................................................................................... ................ 34
2.3.3 Reliability and Robustness ......................................................................................................... 35
2.3.4 Availability................................................................................................................... ............. 37
2.3.5 Maintenance.................................................................................................................... .......... 37
2.3.6 Compatibility with Existing Infrastructure and Vehicle............................................................. 39
2.4 DESIGN ISSUES ............................................................................................................................... ... 40
2.4.1 Vehicle Technologies ................................................................................................................. 40
2.4.2 Infrastructure Design................................................................................................................. 41
2.4.3 Driver Vehicle Interface ( DVI) .................................................................................................. 42
2.5 DEPLOYMENT ISSUES......................................................................................................................... 43
2.5.1 Institutional Issues ..................................................................................................................... 43
2.5.2 Risk Management..................................................................................................................... . 44
2.5.3 Driver Acceptance ..................................................................................................................... 45
2.5.4 Public and Passenger Acceptance .................................................................................................................. 46
2.6 SUMMARY ............................................................................................................................... .......... 47
3.0 ANALYSES OF NEEDS, BENEFITS/ COSTS AND DEPLOYMENT ISSUES............................. 49
3.1 BENEFITS OF LANE ASSIST AND PRECISION DOCKING SYSTEMS......................................................... 49
xiv
xv
3.1.1 Enables Buses to Operate Within Narrow Lanes that Facilitate Dedicated BRT Deployment.. 49
3.1.2 Provides Consistent Travel Time, Especially at Bus Stops ........................................................ 50
3.1.3 Helps to Reduce Overall System Cost........................................................................................ 51
3.1.4 Provide Rail- Like Image and Service......................................................................................... 51
3.1.5 Helps Reduce Driver Stress ....................................................................................................... 51
3.1.6 Helps Ensure Mobility and Accessibility ................................................................................... 52
3.1.7 Enhances Safety ......................................................................................................................... 52
3.1.8 Facilitate Environmentally- Friendly BRT Deployment ............................................................. 52
3.1.9 Facilitate Incremental and Flexible Deployment....................................................................... 52
3.2 ECONOMIC EVALUATION.................................................................................................................... 52
3.2.1 Unit Costs of Electronic Guidance Technologies ...................................................................... 53
3.2.2 Precision Docking Evaluations................................................................................................. 54
3.2.2.1 Lane Transit District ............................................................................................................................... ... 54
3.2.2.2 Los Angeles County Metropolitan Transportation Authority ( LACMTA) ................................................. 56
3.2.2.3 Alameda- Contra Costa Transit District ( AC Transit).................................................................................. 59
3.2.3 Lane Assist Evaluations ............................................................................................................. 61
3.2.3.1 Lane Transit District ............................................................................................................................... ... 61
3.2.3.2 Los Angeles County Metropolitan Transportation Authority ( LACMTA) ................................................. 62
3.3 SAFETY BENEFITS OF LANE- ASSIST AND PRECISION DOCKING SYSTEMS .......................................... 63
3.3.1 Incident Data Analysis for Transit Buses................................................................................... 64
3.3.1.1 Vehicle Collisions by Scenarios.................................................................................................................. 64
3.3.1.2 Passenger Injuries by Bus Maneuver .......................................................................................................... 67
3.3.1.3 Comparison with National Incident Data .................................................................................................... 68
3.3.2 Incident Data Analysis for Light Rail Transit ( LRT) ................................................................. 68
3.3.2.1 LRT Fatalities ............................................................................................................................... ............. 69
3.3.2.2 LRT Severe Incidents...................................................................................................................... ........... 69
3.3.2.3 LRT Collisions..................................................................................................................... ...................... 70
3.3.2.4 LRT General Vehicle Collisions ................................................................................................................. 71
3.3.2.5 LRT Passenger Injury ............................................................................................................................... . 71
3.3.3 Observations .............................................................................................................................. 72
3.4 DEPLOYMENT ISSUES......................................................................................................................... 74
3.5 SUMMARY ............................................................................................................................... .......... 77
4. DEVELOPMENT OF REQUIREMENTS AND SPECIFICATIONS, MEASURES OF
EFFECTIVENESS AND SAFETY ANALYSIS METHODS................................................................. 78
4.1 FUNCTIONAL ANALYSIS OF LANE ASSIST AND PRECISION DOCKING SYSTEMS .................................. 78
4.1.1 Functional Blocks of Lane- Assist Systems ................................................................................. 78
4.1.2 Functional Decomposition......................................................................................................... 79
4.2 PRELIMINARY FUNCTIONAL REQUIREMENTS FOR LANE ASSIST AND PRECISION DOCKING SYSTEMS 81
4.2.1 Performance Requirements........................................................................................................ 81
4.2.1.1 Precision docking performance ................................................................................................................... 81
4.2.1.1.1 Docking accuracy ................................................................................................................... 81
4.1.1.1.2 Transition characteristics ........................................................................................................ 81
4.2.1.2 Lane keeping performance .......................................................................................................................... 82
4.2.1.2.1 Accuracy....................................................................................................................... ......... 82
4.2.1.2.2 Operating conditions - All weather conditions, with transition initiated by drivers .................... 82
4.2.1.2.3 Transition characteristics ........................................................................................................ 82
4.2.2 Safety......................................................................................................................... ................ 82
4.2.3 Reliability.................................................................................................................... .............. 83
4.2.4 Availability................................................................................................................... ............. 83
4.2.5 Maintainability................................................................................................................ .......... 84
4.2.6 Compatibility with Existing Infrastructure ................................................................................ 85
4.3 PRELIMINARY TECHNICAL SPECIFICATIONS....................................................................................... 85
4.3.1 Vehicle position sensing capability ........................................................................................ 85
4.3.1.1 Spatial Coverage....................................................................................................................... ............. 86
4.3.1.2 Resolution..................................................................................................................... ......................... 86
4.3.1.3 Robustness with respect to environmental changes ................................................................................ 86
xvi
xvii
4.3.1.4 Timing and update rate ........................................................................................................................... 86
4.3.2 Subject vehicle status sensing capability................................................................................ 86
4.3.2.1 Vehicle status parameters..................................................................................................................... ...... 86
4.3.2.2 Events......................................................................................................................... ............................... 87
4.3.3 Actuation...................................................................................................................... ............ 87
4.3.3.1 Steering actuator ............................................................................................................................... ..... 87
4.3.3.2 Engine and brake actuator....................................................................................................................... 88
4.3.4 Driver- Vehicle Interface system................................................................................................ 88
4.3.4.1 Interface Contents ............................................................................................................................... ....... 88
4.3.4.2 Processing capability..................................................................................................................... ............. 89
4.3.4.3 Display update time........................................................................................................................... ......... 89
4.3.5 Infrastructure requirements ...................................................................................................... 89
4.3.5.1 Roadway sensing and construction ........................................................................................................... 90
4.3.5.2 Communication capability ......................................................................................................................... 90
4.3.6 Driver qualification and training requirements......................................................................... 90
4.3.6.1 Qualification ............................................................................................................................... ............... 90
4.3.6.2 Training ............................................................................................................................... .................. 91
4.3.7 Maintenance interval minimum requirements............................................................................ 91
4.4 MEASURES OF EFFECTIVENESS FOR EVALUATING LANE ASSIST SYSTEMS ........................................ 91
4.4.1 Incremental costs to provide lane- assist capabilities................................................................ 92
4.4.2 Operational improvements......................................................................................................... 92
4.4.2.1 Travel Time Saving ............................................................................................................................... 92
4.4.2.2 Time Savings at Bus Station ................................................................................................................... 93
4.4.2.3 Time Savings for Passenger Boarding and Alighting ............................................................................. 93
4.4.2.4 Passenger Ride Quality........................................................................................................................ .. 93
4.4.2.5 Operating Speed.......................................................................................................................... ........... 93
4.4.2.6 Lateral Position Accuracy....................................................................................................................... 93
4.4.3 Safety......................................................................................................................... ................ 93
4.4.4 System Reliability and Availability ............................................................................................ 94
4.4.4.1 Mean time between failures ( MTBF)................................................................................... 94
4.4.4.2 Availability .......................................................................................................................... 94
4.4.5 Maintenance Burden.................................................................................................................. 94
4.4.6 Operational Limitations............................................................................................................. 95
4.4.7 Infrastructure Effects ................................................................................................................. 95
4.4.8 Public Perception..................................................................................................................... . 95
4.4.9 Driver Acceptance ..................................................................................................................... 96
4.5 SAFETY ANALYSIS ....................................................................................................................... 96
4.5.1 Determination of Safety Integrity Level ..................................................................................... 96
4.5.2 Preliminary Hazard Analysis of Lane Assist Systems ................................................................ 97
4.5.2.1 Categories of Hazards in Lane Assist Systems............................................................................................ 97
4.5.2.2 Hazard Analysis for Lane Assist and Precision Docking Systems.............................................................. 98
4.5.2.2.1 Hazard Identification................................................................................................................. . 98
4.5 .2.2.2 Relationship Between Failures and Hazards .............................................................................. 99
4.5.2.3 Failure Mode Effect and Criticality Analysis............................................................................... 100
4.6 SUMMARY ............................................................................................................................... ........ 104
5.0 CONCLUSION..................................................................................................................... ............. 105
REFERENCES..................................................................................................................... ................... 107
APPENDIX A – EFFECTS OF TIGHT TURNING RADII ON NEEDED LANE WIDTH............. 110
APPENDIX B: QUESTIONS FOR LANE ASSIST REQUIREMENTS WORKSHOP.................... 114
xviii
xix
LIST OF FIGURES
Figure 1.1 Mechanical Guide wheel on bus in Essen, Germany………………………. 3
Figure 1.2 Las Vegas CiViS bus with vision based guidance………………………….. 6
Figure 2.1 LTD Proposed BRT Corridors……………………………………………… 22
Figure 3.1 Vehicle Collisions by Collision Scenario in Five Fiscal Years……………... 66
Figure 3.2: Passenger Injuries by Bus Maneuver in Five Fiscal Years…………………. 68
Figure 3.3 General LRT Collision Incidents by Initial Point Of Impact………………... 70
Figure 4.1 Functional Block Diagram of Lane Assist and Precision Docking Systems... 78
Figure 4.3 Safety Integrity Level ( SIL) Determination………………………………… 97
Figure A1 - Additional lane width required vs turning radius
for a 40 ft New Flyer bus……………………………………………… 110
Figure A2 - Additional lane width required vs turning radius
for a 60 ft New Flyer articulated bus……………………………………. 111
Figure A3 - Vehicle motion………………………………………………………….. 111
Figure A4 - Turning radius for single unit bus………………………………………. 112
Figure A5 Offset at rear tire ( m)……………………………………………………… 112
Figure A6 - Turning radius of articulated bus……………………………………….. 113
Figure A7 - Offset at rear tire ( m)……………………………………………………. 113
xx
xxi
List of Tables
Table 1.1 - Summary of Infrastructure Characteristics for Various Lane Assist
and Precision Guidance Systems……………….………………………. 11
Table 1.2 - Summary of Vehicle Characteristics for Various Lane Assist
and Precision Guidance Systems……………………………………….. 12
Table 1.3 Bus Operator Health Problems ( Evans ( 1999)……………………………. 13
Table 1.4 Temporal distribution of gaze directions ( from Gobel et al 1998)………... 14
Table 3.1 Projected Unit Costs of Lane Assist System……………………………… 53
Table 3.2: Break- Even Bus Docking Time Savings for LTD……………………….. 55
Table 3.3 Precision Docking Sensitivity Analysis for LTD………………………… 56
Table 3.4 Break- Even Bus Docking Time Savings for LACMTA…………………… 58
Table 3.5 Precision Docking Sensitivity Analysis for LACMTA…………………… 59
Table 3.6 Break- Even Bus Docking Time Savings for AC Transit………………….. 60
Table 3.7: Precision Docking Sensitivity Analysis for AC Transit………………….. 61
Table 3.8 Collisions with Severe Injuries……………………………………………. 65
Table 3.9 Collision Scenarios……………….……………………………………….. 66
Table 3.10 Passenger Injury Bus Maneuvers………………………………………… 67
Table 3.11 Incident and Cost Data For Light Rail ( 5 Years)…………………………. 69
Table 3.12 LRT Fatalities in Five Fiscal Years………………………………………. 69
Table 3.13 Severe LRT Incidents in Five Fiscal Years……………………………… 70
Table 3.14 LRT Intersection Collisions in Five Fiscal Years………………………… 71
Table 3.15 Passenger Injuries in Five Fiscal Years…….…………………………….. 72
Table 4.1 Relationship Between Failures And Hazards ……………………………... 100
Table 4.2 FMECA for Lane Assist and Precision Docking System…………………. 101
xxii
.
1
1.0 INTRODUCTION
Bus Rapid Transit ( BRT) systems can provide high quality, high capacity bus transit
service on easily identifiable route structures at a cost lower than that of urban light rail
systems. They may apply integrated land use planning and advanced design concepts, as
well as intelligent transportation systems ( ITS) concepts and technologies, to provide
significantly higher operating speeds, greater service reliability, and increased
convenience. BRT can thereby become a cost- effective alternative to urban light rail
systems, with the potential to attract non- traditional riders and contribute to the reduction
of traffic congestion. Many regions in the United States are very interested in planning
for and deploying BRT. Although each of the BRT deployment sites has unique features,
all BRT- interested agencies are commonly interested in applying advanced operational
concepts and technologies, particularly Intelligent Transportation Systems ( ITS)
technologies that can improve efficiency, safety and service level of BRT operation. BRT
interested agencies have expressed strong interest in incorporating electronic guidance
technologies.
Electronic guidance can be applied to transit buses to provide lane assist and precision
docking functions. Lane assist allows the bus to operate in a designated lane that is only
inches wider than the bus itself, while precision docking enables buses to precisely stop
at bus stations without increasing driver workload. Electronic guidance can be
implemented with partially or fully- automated modes to guide buses through narrow
bridges, tunnels, toll booths, and roadways, as well as bus stops, tight curves, and
designated trajectories in maintenance yards. The precision docking capability at bus
stops, which allows fast loading and unloading of passengers with special needs, thereby
reduces station dwell time.
Electronic guidance technologies provide numerous ‘ rail- like’ features that enhance
efficiency, safety and quality of service for the BRT operations. They may become
critical components of a BRT design when the space for the planned BRT lane is
constrained. Electronic guidance may also be considered as an option to augment the
initial BRT design, where this option is not essential but could provide significant
performance improvements. Lane assist systems can be implemented in mixed traffic
lanes, which the bus shares with normal traffic, or in dedicated bus- only lanes, which
could be separated from the other lanes by road markings or by physical structures
( barriers).
The deployment of lane assist systems can be viewed from three perspectives: that of the
transit agency, the transit driver, and the transit passenger. For the agency, lane assist
systems offer significant benefits including the delivery of rail- like service, an attractive
feature to riders, at a fraction of rail cost. From the driver’s perspective, the lane assist
system can be a means to decrease workload and stress while at the same time allowing
him/ her to operate in more challenging environments ( e. g., narrower lanes). For
passengers, the implementation of an electronic guidance system will mean smoother
2
operation, faster and safer boarding and alighting, better schedule reliability, and
increased mobility for ADA riders.
1.1 Project Background
Under the sponsorship of FTA, the Minneapolis- Saint Paul Metropolitan Transit Agency
( Metro) and University of Minnesota ITS Institute developed requirement specifications
for a lane assist system. FTA also requested a separate effort led by the California
Department of Transportation ( Caltrans) to develop inputs on system needs and
requirements. In response to FTA’s request, AC Transit, Los Angeles County
Metropolitan Transportation Authority ( LACMTA), Lane County Transit District ( LTD),
and San Diego Association of Governments ( Sandag), the California Department of
Transportation ( Caltrans), Gillig Corporation and California PATH Program ( PATH)
formed a partnership to support and supplement the Metro team’s work in developing
requirements and specifications for transit lane assist and precision docking systems.
System requirements stem from stakeholder needs. AC Transit, LACMTA, and LTD are
all members of the BRT Consortium. These agencies have planned dedicated BRT routes
and are convinced that electronic guidance technologies can offer benefits in enhancing
the efficiency, safety and quality of BRT service. This report examines the needs for and
potential applications of lane assist and precision docking systems for these BRT sites.
Through the case studies, this project has ( 1) defined transit agency needs for lane assist
technologies and ( 2) defined both performance requirements and technical specifications
for lane assist and precision docking systems. As such, the emphasis of this project has
been placed on receiving inputs from partner transit agencies through workshops and
close interactions. The case studies of the BRT sites have addressed the following issues:
o Needs, functionalities, and applications of lane assist systems
o Cost- benefit analysis of lane assist systems and their specific types of benefits
( and potential synergies with other advanced technologies)
o Drivers’ perspectives on lane assist systems
o Operation environment and conditions, including constraints
o Maintenance aspects of lane assist systems.
The synthesized inputs from the transit agencies captured the common and special needs,
benefits and constraints of each type of application for the transit lane assist and precision
docking systems. Based on the stakeholders’ inputs, in- depth studies on benefits and
costs have been conducted. The stakeholder perspectives also provided qualitative
requirements, which have been translated into system definitions, performance
requirements and technical specifications.
1.2 Technology Options
Electronic guidance technologies are designed to aid the driver in controlling vehicle
position within designated bus lanes ( lane keeping or lane assist) or to follow a specific
trajectory when coming to a stop ( precision docking). When engaged, lane assist systems
3
are designed to perform these functions without driver input, although they may also be
designed to provide “ assist” functions that augment a driver’s steering actions to provide
more accurate lane keeping performance. In either case, systems applying electronic
guidance technologies are necessary to maintain safety because of the smaller margin for
error associated with operating a bus in narrower lanes, especially adjacent to regular
traffic.
Assisted guidance has been in development for the past thirty years. What differentiates
the various technologies is their means of position sensing. Most have been thoroughly
tested in the research phase and their advantages and disadvantages assessed. The
appropriateness of each of these technologies is likely to depend heavily on the
circumstances associated with the particular BRT operation being considered.
Lane assist systems can be classified as either mechanical or electronic, depending on
how the lane deviation and correction are performed. The Metro team has conducted a
technology review and the results were reported in [ Donath et al., 20031]. Under this
project, a scanning tour with participation of FTA, FHWA, transit agencies and PATH
was conducted to collect information from several European organizations that have had
experience in the development and operation of transit lane assist systems based on three
different technologies, including ( 1) optical guidance in Rouen, France, ( 2) magnetic
guidance in Eindhoven, Netherlands and ( 3) mechanical guidance in Essen,
Germany. Findings were documented in [ Shladover, et al, 20052]. The following
provides a high- level overview of the electronic guidance technologies reviewed by these
two studies and a summary of the advantages and disadvantages of each technology.
1.2.1 Mechanical Guidance
Mechanical guidance relies on physical contact
between the vehicle and the infrastructure. Location
of the vehicle relative to the roadway is determined
via changes in the nature of this physical contact. An
example of this type of system is used in Essen,
Germany, where each bus is fitted with special “ guide
wheels” by the front wheels, allowing the bus to
transfer from a road onto the track in one smooth,
easy movement. The guide wheels are directly
connected to the vehicle’s steering mechanism and
once these guide wheels are locked in place, the track
is effectively steering the bus. Another type ( e. g., in
Nancy, France) utilizes central guide wheels which are
in contact with an underground guide rail in a slot in the middle of the bus lane.
1 Adapted from Minneapolis Metro Transit and the University of Minnesota ITS Institute ( February 2003), Bus Rapid
Transit Lane Assist Technology Systems Volume 1: Technology Assessment, Minneapolis, MN
2 California PATH, et. al., Lane Assist Systems in Europe: Report on Technical Visit to Europe on Transit Lane Assist
Technologies
Figure 1.1 Mechanical Guide
wheel on bus in Essen, Germany
4
Advantages of mechanical guidance technologies
• High reliability – often very low tech, direct mechanical linkages can be employed to
ensure high reliability under all kinds of conditions.
• Very high positional accuracy.
• Relatively low per vehicle cost – mechanical sensing systems may be less expensive
than electro- optical sensing.
• Because the systems are mechanical with exposed parts, visual and manual inspection
is straightforward and can be performed at the beginning of each driver’s shift.
• These systems are relatively insensitive to weather and other environmental
conditions; however, ice jamming of the buried guidance rail can be a problem.
• Finally, in the ( rare) event of a system failure, a busway, with vertical curbs provides
a physical means with which to keep the bus out of adjacent lanes, keeping the
passengers and motorists in the adjacent lanes safe.
Disadvantages of mechanical guidance systems
• Often requires a dedicated lane leading to potentially high infrastructure modification
costs. This will depend heavily on the type of infrastructure required and the
infrastructure already in place or planned for the BRT site. A transit agency whose
right- of– way is extremely limited may be forced to eliminate this system from
consideration.
• Potential impediments to other vehicle traffic. Some types of infrastructure
modifications to support mechanical guidance, ( e. g. rails) may preclude the use of the
lanes by other traffic. In a dedicated lane scenario this should not be a problem. This
system would negate use of lanes by emergency vehicles.
• Sensitivity to vehicle failure – A breakdown of a bus on a narrow, hard- barrier
separated narrow lane could result in a system- wide shutdown since the disabled bus
will not be easy to move. Diverting other automated buses around the bottleneck
would be difficult since lane “ jumping” is precluded by the barriers.
• It is difficult to design a fixed feedback mechanical mechanism that can meet
performance requirements under different operating conditions. Transition between
guided and unguided areas, for example, may not be smooth if not properly designed
( http:// www. lightrailnow. org/ features/ f_ ncy001. htm)
• The tight physical tolerances on the construction of the mechanical guiderails lead to
high infrastructure costs, virtually as high as for light rail in the case of Essen.
1.2.2 Electronic Guidance
Electronic guidance systems use sensors to detect a vehicle’s lateral deviation from the
lane center. This deviation is then fed back to a computer where a corrective steering
command is calculated and sent to a steering actuator which will steer the vehicle back to
the center in order to maintain the vehicle within lane boundaries. Determining the
vehicle’s lateral deviation with high accuracy, high bandwidth and robustness is very
important to the successful implementation of electronic guidance.
5
1.2.2.1 Vision- based Guidance
Vision- based guidance systems use optical methods to determine the vehicle’s location
relative to the lane by the sensing of features such as lane markings. This optically
derived information is used to provide lateral control assistance to the driver. Optical
sensors can vary in their sensing technology ( passive vs. active) and their orientation
( forward looking vs. downward looking). Optical sensors include traditional video
cameras as well as infrared ( IR) cameras.
General advantages of optical technologies
Typically fewer infrastructure modifications are required since they can sometimes work
with existing lane markings or other road features. Because of their flexibility and
relatively minor infrastructure needs, optical technologies can provide benefits beyond
the specific narrow lanes where control assistance for the driver is required. This will be
particularly useful in BRT scenarios where buses need to leave the narrow designated
travel lanes and travel on normal roads. This extended functionality could include lane
assistance similar to that available on the narrow lanes or passive warnings/ advisories
directed toward the driver. Such flexibility will allow for easy diversion from a typical
route if conditions require ( e. g. breakdown of another bus on narrow lane).
Optical technologies may also be employed to support other functions that may be
important for a BRT system, including precision docking, longitudinal ( headway)
control, pedestrian detection, and sign recognition. Other advantages include:
• Position accuracy – This is particularly possible with downward looking passive or
active systems.
• Small size – Optical sensors are usually small, sometimes down to one cubic inch.
This makes packaging, installation and maintenance relatively easy, and does not
detract from the appearance of the bus.
• While components will tend to be higher cost, the simplicity of the hardware will
make moving optical technologies from one bus to another relatively easy, potentially
mitigating the long- range cost issue.
• Less chance of sensor interference with other systems both on and off the bus.
Disadvantages of optical technologies
• Typically increased sensitivity to environmental factors such as lighting, weather or
pavement conditions. This is particularly true with passive vision- based guidance
systems.
• Potentially higher per vehicle cost as the sensors involved ( be they cameras, lasers or
radar) are likely to be more expensive than some of the other technologies.
• If the pavement markings used by the optical system are not consistent with normal
roadway markings, they are likely to cause confusion for the drivers of other vehicles
that would share lanes with the guided buses.
• Temperature sensitivity – Typical off- the- shelf consumer/ commercial cameras may
not operate at the high temperatures outside a bus operating in an area such as Florida
6
during the summer months. This may require use of industrial/ military cameras,
and/ or environmental management strategies which could add to the size and cost of
system.
Even in light of advances in image processing software and the computer hardware on
which it runs, the applicability of vision- based systems is limited to relatively structured
environments for which good atmospheric conditions exist. This is evidenced by the Las
Vegas CiViS system. For its operation, the CiViS needs a specific pattern to recognize,
and it needs a clear view of that pattern. Frequent roadway maintenance may be needed
to maintain sufficient visibility of the optical pattern, which has been a significant
problem with dust, dirt and melting roadway asphalt in Las Vegas. Areas in which snow,
heavy rain, and fog are endemic cannot reliably be serviced by a vision- based system. As
these conditions are faced by a significant number of U. S. transit agencies, vision systems
should be ruled out as a primary electronic guidance mechanism for these transit
agencies.
Figure 1.2 Las Vegas CiViS bus with vision based guidance
It is also important to note that repainting of lines is not an inexpensive proposition; in
fact, Las Vegas intended to use the CiViS guidance system only for precision docking,
and not for lane assist. This decision was based on the fact that painted stripes last only
for a short period of time when exposed to the high heat and intense UV rays of the
desert. It would have been too expensive to periodically repaint the lane markers over the
entirety of the bus route.
1.2.2.2 Magnetic Guidance
Magnetic guidance systems use magnetic material ( e. g., magnetic tape or discrete plugs)
located on, adjacent to, or embedded in the roadway. Sensors of the magnetic field
onboard the vehicle are used to determine the vehicle’s position relative to the lane and to
provide lane keeping assistance to the driver.
7
Advantages of magnetic technologies
• Insensitive to environmental factors such as lighting, weather and pavement
conditions.
• Very high position accuracy possible.
• Static coding of other information is possible – for example: warning of upcoming
road curvature by varying the polarity of discrete magnets in a known pattern.
Disadvantages of magnetic technologies
• Sensitivity to other ferromagnetic material in the vicinity of the bus such as
components in the vehicle, roadway structural supports or reinforcing rebar may
distort the magnetic field. Such changes in the background magnetic field are
sometimes hard to isolate and can deteriorate system performance.
• The low field strength provided by the in- road magnets limits the maximum range for
which the lateral position can be reliably estimated.
• Requires some modifications to the infrastructure. This will not be a substantial
impediment when there are only a limited number of lane- miles where lane- keeping
assistance is required.
The California PATH Program has performed extensive experimentation and
development of magnetic marker based lateral control/ guidance system on different
vehicles. In one such example, a supplemental guidance display was installed in a
California Department of Transportation ( Caltrans) snowplow in order to improve the
safety and efficiency of snow removal operations. Lane position information was
calculated based on the magnetic markers embedded in the roadway and “ read” by a
single magnetometer array comprised of seven magnetic sensors installed at the front of
the snowplow. Signal processing of the magnetometers provides lateral position
measurement relative to the center of the lane, longitudinal position relative to mileposts,
and yaw angle estimate. Binary coding of the magnetic markers when installed ( north
pole up vs. south pole up) also can provide information about upcoming roadway
characteristics, e. g. the direction and radius of the curves ( Tan et al., 2001; Zhang et al., ).
During the period of October 6 to December 1, 2000, a magnetic guidance equipped
Buick LeSabre underwent a series of tests on the test track of the Public Works Research
Institute ( PWRI) in Tsukuba City, Japan. The installed system provided information to
either an automatic steering command to the steering actuator, or a display providing a
preview of the future vehicle position ( predictor) if the driver does not correct his steering
action. The design of this guidance display was optimized to make it very easy for the
driver to steer the vehicle accurately, even in zero- visibility conditions. Furthermore,
PATH also demonstrated a smooth switching method that was previously developed,
allowing the driver to change between automatic and manual steering control at any
location or time that he commands ( Tan and Bougler, 2001).
Recently, two 40 foot CNG New Flyer buses and one 60 foot articulated New Flyer bus
were retrofitted with magnetic maker based lateral control system. Precision inline and s-
8
curve docking and stopping maneuvers of the 40 foot bus were successfully demonstrated
in Washington DC and San Diego with 2 cm accuracy laterally and 10 cm accuracy
longitudinally. Lane assist, lane change and automated/ manual transitions were
demonstrated on the I- 15 HOV lanes in San Diego, achieving 15 cm lateral tracking
accuracy at up to 65 mph for both the 40 foot single unit bus and 60 ft articulated bus.
Toyota is developing an Intelligent Multimode Transit System ( IMTS) which uses
magnetic markers as the lateral sensing system. Experimental studies on a test course
show ± 5 cm lateral deviation at up to 30 km/ h ( Aoki and Suyama, 2000). In 2005,
Toyota demonstrated their IMTS system at the Aichi World Exposition near Nagoya,
Japan, with platooning, lane assist and precision docking functions, and carried more than
10 million passengers on the IMTS buses..
The Phileas bus has an electronic lane assistance and precision docking system with all-wheel
steering. The system is based on magnetic markers every 4 meters in the road
surface and works at speeds up to 80 km/ h and under most weather conditions. When
driving in automatic mode, the Phileas bus requires only 6.4 m of width for two- way
dedicated lanes at 70 km/ h ( http:// www. apts- phileas. com).
1.2.2.3 Wire Guidance
Wire guidance systems share many similarities with magnetic guidance systems, but in
this case, an electrified wire is buried in the pavement and its position relative to the
vehicle is sensed. While it shares the same advantages and disadvantages as magnetic
guidance, it has an additional disadvantage. The electric current required for this type of
system to operate will be lost if the wire is broken, causing a single point of failure
rendering the system inoperable. This could be a particular problem in northern climates
where deflections in the pavement due to temperature variations may cause wire
breakage. This introduces a liability into the system which is beyond the control of the
operating agency.
1.2.2.4 Global Positioning System ( GPS) Guidance
GPS guidance systems use the constellation of satellites maintained by the U. S.
government and special receivers to localize the vehicle on a digital map of the
environment. This information can be used to determine the vehicle’s position relative to
the lane.
Advantages of GPS technologies include:
• Little infrastructure modifications are required. To achieve the kind of accuracy
required for electronic guidance, a GPS system may require the installation of a base
station for differential corrections, but the cost of such a base station is relatively low.
• Relatively high positional accuracy is possible.
9
• Because of their flexibility and independence from the local infrastructure, GPS
technologies can provide benefits beyond the specific narrow lanes where control
assistance for the driver is required. This would be particularly useful in BRT
scenarios where the buses need to leave the narrow, designated travel lanes and travel
on normal roads. This extended functionality could include lane assist similar to that
available on the narrow lanes or passive warnings/ advisories directed at the driver.
• Dual use – In addition to their use in lane assistance, GPS technologies can be
employed to support other functions that may be important for a BRT system
including precision docking, longitudinal ( headway) control, vehicle routing,
scheduling, etc.
• Systems are relatively immune to dynamic environmental influences. GPS systems
are not degraded by weather, lighting or pavement conditions.
Disadvantages of GPS technologies include:
• Relatively high per vehicle cost. GPS sensors accurate enough to provide useful lane
position data are relatively expensive compared with the other technologies.
• Sensitive to static environmental factors such as occlusion of sky by trees, bridges,
overhead signs, nearby buildings, hills etc., resulting in significantly degraded
performance and require additional expensive sensors ( e. g. inertial sensors) to fill in
the “ gaps”.
• For differential GPS in which the bus receives GPS- correction radio signals
transmitted from a private service provider, excellent positioning accuracy can be
achieved; however the service provider controls the differential broadcasts, which
introduces a potential single point of failure beyond the transit agency’s control.
• GPS signals, being very weak, are vulnerable to interference and “ jamming”, which
can create concerns about their reliability and availability for such a safety- critical
application.
California PATH has conducted several projects dealing with improvement of GPS
technology and vehicle control/ guidance with GPS based positioning system. Carrier
phase signal processing and DGPS/ INS ( inertial navigation system) integration were
investigated to see if they could overcome problems associated with GPS based
positioning systems such as accuracy, latency and low updating rate. Researchers found
that the integrated CP ( Carrier- Phase) DGPS/ INS system could provide vehicle position,
velocity, acceleration, heading and angular rate at 150 Hz with accuracies ( standard
deviation) of 1.5 cm, 0.8 cm/ s, 2.2 cm/ s/ s, 0.1 deg and 0.1 deg/ s respectively ( Farrell and
Barth, 2002). A CP DGPS/ INS based control system was tested onboard a PATH vehicle
at the Crow’s Landing test facility. Decimeter accuracy was achieved up to 70 mph
under open sky conditions ( Farrell et al., 2003; Tan et al., 2003).
In Minnesota, research was conducted in which CP DGPS was used for snowplow
guidance. Integration with INS was used to address the GPS signal loss due to certain
intermittent blockage from bridges and canyons. If the signal loss lasted less than 30
seconds, estimation from INS was used for guidance. For signal losses greater than 30
10
seconds, magnetic tape embedded in the roadway was used to provide lateral position
( Minnesota DOT, 2001; Minnesota DOT, 2002).
In the Minneapolis/ St. Paul Metro Area, a lane support system retrofitted on a Metro
Transit bus was demonstrated to be capable of steering a 9.5 ft wide bus along a 10 ft
wide “ bus only shoulder.” Two CP DGPS receivers ( Trimble ms 750) were used to
provide centimeter accurate position, roll and heading information. Thirteen cm
( standard deviation) lateral tracking error was achieved at speeds up to 35 mph ( Donath
et al., 2003).
Most GPS based control/ guidance research is carried out in an ideal or a semi- ideal
environment where sufficient satellites are available with little problem of blockage and
multi- path error. In reality, such an environment does not always exist. Urban canyons,
nearby buildings, bridges, overpasses, tunnels, roadside trees, even a heavy vehicle in the
next lane will occur with high statistical certainty. The always- changing satellite
configuration makes the reliable prediction of such degraded characteristics virtually
impossible in many cases. Accuracy may deteriorate or be lost for significant periods of
time.
1.2.3 Comparison of different technologies
The Metro team summarized the infrastructure and vehicle characteristics of different
electronic guidance technologies using two tables ( Donath et al., 2003). Based on the
findings from the European study visit and PATH’s experience with guidance
technologies, this table has been updated. Note that when reviewing different
technologies for use in specific BRT applications, in addition to the characteristics listed
in the table, the following additional key points also need to be considered:
• Technology and infrastructure need to be compatible with the types of weather and
road conditions that may be encountered, including bright sun, fog, snow, ice, heavy
rainfall, strong wind, high humidity, and extreme temperatures. External hardware
will be required to withstand dust and water as well as sand and salt.
• A system failure should not prevent the bus from operating under driver control
within the BRT infrastructure ( although this may be at reduced speeds) or off route.
• No single point of failure should be able to jeopardize the availability or the operation
of the system. This is especially important with aspects beyond the control of the
operating agency, such as GPS differential correction beacons.
11
TABLE 1.1 - Summary of Infrastructure Characteristics for Various Lane Assist and Precision Guidance Systems
Technology Production
Status
Road
Infrastructure
Cost/ Mile
Supporting
Infrastructure
Costs
Dedicated lane Weather
Limitations
Topographical
Limitations
Curb Guidance Presently out of
production
$ 2.65M / mile 0 Yes Heavy snow & ice
problematic
None
Rail Guidance Prototype ( 2
systems)
$ 15.5 M / mile 0 No Ice may jam up
guide rail
None
Vision Guidance In Production None Cost of surveying,
painting and
repainting
reference stripes
No Yes – fog, heavy
rain, snow in air,
or on ground, UV
& heat on paint
stripes
Some – roads
must be kept clear
so stripes are
visible.
Discrete Magnets PATH Prototype None $ 10,000 mile
( survey &
installation of
magnets)
No No None
3M Magnetic
Tape
No Longer
Supported
None $ 3 - $ 5 per linear
foot of magnetic
tape, installed
No No None
DGPS University of
Minnesota
Prototype ( one
system on one
bus)
None $ 250 / lane- mile
to map roadway,
GPS base stations
at $ 25 K each +
base station
software
~$ 100,000
No No Yes – need clear
view to sky for
satellite signals
Source: Bus Rapid Transit Lane Assist Technology Systems, Volume 1 [ with a few modifications]
12
TABLE 1.2 - Summary of Vehicle Characteristics for Various Lane Assist and Precision Guidance Systems
Technology Vehicle sensor cost Computational
Complexity
Lane
Assist/ Precision
Docking
Control Features Bus Features
Curb Guidance $ 15,000 - $ 30,000 None Yes/ Yes Mechanically
actuated steering
system
Conventional bus
equipped with
mechanism
Rail Guidance Not Known Low Yes/ Yes Mechanical or
Hydraulic
connection to guide
rail
Low floors, Euro
design, 3 articulated
sections
Vision Guidance
( CiViS)
Vehicle cost is ~$ 1
M per vehicle,
estimate 10% is
technology cost
High Yes / Yes Electrically
actuated steering
system
CiViS – Low floors,
Euro styling
Discrete Magnets
( PATH)
$ 5000-$ 10,000 for
sensors,
Medium Yes/ Yes Electrically actuated
steering system,
retrofit
Retrofit onto
existing bus
3M Magnetic Tape $ 5000-$ 10,000 for
sensors,
Medium Yes/ Yes
( modifications
needed for low
speeds)
Electric steering,
retrofit
Retrofit onto
existing bus
DGPS ( University
of Minnesota)
$ 25,000 - $ 30,000
for sensors ( in
volume)
Medium Yes/ Yes Electric steering,
retrofit
Retrofit onto
existing bus
Source: Bus Rapid Transit Lane Assist Technology Systems, Volume 1
1.3 Human Factors Considerations
Electronic guidance systems for transit lane assist and precision docking applications are
intended to enhance the performance of transit buses, thereby to improve service quality.
On the other hand, these technologies must be designed to be easily adopted by bus
drivers and will reduce, rather than increase, their stress level and improve the safety.
Previous research has shown that urban bus driving is a stressful occupation that can lead
to long- term health difficulties. While there have not been any long- term studies of an
implementation of lane assistance systems, initial short trial studies have had some
promising results, suggesting that such a system may aid a decrease in the stressful nature
of urban bus driving. There is however a clear need for future research to investigate
how driving with an electronic guidance system affects subsequent driving when driving
in a manual mode, as well as issues such as how well humans can perform as the monitor
of such a system and how well drivers, once trained, can take control if a fault does
occur.
1.3.1 Bus Operation is a Stressful Vocation
It has been suggested that urban bus operation is a very stressful vocation ( Evans et al
1999). In a review of a number of previous studies, Evans ( 1999) concludes that bus
operators have higher rates of the conditions listed in Table 1.3 compared with people
from similar occupations:
Table 1.3 Bus Operator Health Problems ( Evans ( 1999)
Condition Urban bus operators compared with people
from similar occupations
Cardiovascular disease Increased levels
Gastrointestinal disease Increased levels
Driver absenteeism rates from illness
typically related to stress related causes
Increased levels
Raised blood pressure Increased levels
Neuroendocrine stress hormones Have elevated levels
Given the high levels of stress related illness much research has been conducted to
determine why urban bus driving is stressful. Meijman et al ( 1998) suggest that bus
operators have three main psychosocial demands in their job; maintaining schedule,
giving good customer service and operating the bus safely. As can be imagined the
above demands often compete, which can cause stress for a bus operator. In addition to
these three main stressors, Evans et al ( 1999) lists from the literature the following
physical stressors evident in the bus operator’s task: dealing with traffic congestion, and
ergonomic factors related to bus operation such as noise, and climatic conditions ( thermal
and air quality).
In order to gain a better understanding of the urban bus driver’s task, Gobel et al, 1998
performed an eye- movement analysis of German bus operators. The total breakdown of
visual scanning patterns can be seen in Table 1.4 below:
14
Table 1.4 Temporal distribution of gaze directions ( from Gobel et al 1998).
Gaze direction Percentage
Outside 73.2
Mirrors 10.8
Window jambs 8.4
Instruments 3.2
Customer service objects 5.0
One of the concerns about adding electronic guidance is that monitoring it should not
take the driver’s time away from monitoring the forward view. To date there has been no
study that we are aware of that has compared drivers’ visual practice with and without
electronic guidance. This issue however does provide us with a design specification that
an electronic guidance system should not require any more attention than other
instruments.
1.3.2 Previous efforts to make urban bus operation less stressful
In a study by Grosbrink et al ( 1998) ways to decrease the physical stressors encountered
in bus operation were investigated. Though a review of ergonomic solutions to physical
stressors is out of the scope of this document, interested readers should review Grosbrink
et al ( 1998).
Evans et al ( 1999) sought ways in which to decrease the stressful nature of urban bus
driving. The researchers implemented changes in Stockholm, Sweden and compared
feedback from drivers operating on “ improved routes” with drivers operating on similar
routes with no “ improvements”. The “ improvements” included; construction of a
separate bus lane for the most congested sections of the routes, changes in the routes to
minimize difficult turns and bottlenecks, the construction of “ passenger peninsulas”
where possible to bring the passengers out to the bus to avoid pull- overs to the curb, the
installation of bus traffic signal priority system and installation of an electronic bus
information system for passengers. As all the implementations were done at one time, it
is not possible to determine the effect of each change individually; however, the
researchers report that based on driver questionnaires, observations, and psycho-physiological
measures the changes did reduce occupational stress among drivers.
Interestingly, many of these interventions are somewhat similar to many of the changes
proposed in current BRT plans. More specifically, lane assist will remove the need for
drivers to pull over to the curb and should also minimize difficult turns.
1.3.3 Previous Lane Assist Human Factors Research
Ward et al, 2003 conducted research to determine the impact of using a lane assist system
developed to provide a “ vehicle control coping support function” when operating a bus
on narrow dedicated highway shoulders. The researchers had bus operators drive under
15
the following three conditions: in a lane adjacent to the narrow shoulder lane manually,
on the shoulder lane manually, and on the shoulder lane with the lane assist system on.
The results from this study suggest that while their experimental system did not reduce
subjective measures of stress for drivers it did improve the stability and control of the
vehicle. The authors concluded that the reason that the subjective stress levels did not
decrease could be attributed to the perceived unreliability of their system and the need for
drivers to interpret what the feedback that the system was giving them meant.
In order to test how bus operators’ workload is affected by an automated docking system
Collett et al ( 2003) measured electrodermal activity ( as a measure of workload) both with
and without an automated docking system during 5 different docking scenarios. The
promising results from this study suggest that while drivers’ workload with a new system
did initially go up, it was reduced as the drivers became more experienced with the
system. The authors of this study also simulated failures of the system that they had not
trained drivers for, which not surprisingly resulted in increased measures of workload. It
would have been interesting to see what effect a failure would have had on workload
measures if the bus operator had been trained for the event. The researchers in this study
found that docking with the system was about 5 cm more accurate both at the front and in
the middle of the bus, closer than when the drivers performed the same docking
manually. Survey responses from the drivers also suggested that the system was useful
and easy to monitor, facilitated docking and thus decreased operator workload. In their
summary Collett et al ( 2003) stressed that for technologies to improve the safety of bus
operation, operators must learn how these technologies work.
To further emphasize the importance of operators fully understanding how technologies
work and when to intervene, Sheridan ( 2002) [ p. 30] cites the example of a fatal
Washington DC Metro accident which was attributed to an unclear management policy
regarding when a driver could take over in case of automation failure:
“ The system had been set in automatic control by the operations control center,
even though some operators had requested a manual control mode because of icy
track. However, one new operator, when he found his train was not slowing as
expected, was intimidated by what he perceived to be orders not to countermand
the automated braking system. His train overran the Shady Grove Station by 470
feet ( about 143 m) and struck another train at full speed.”
As electronic guidance technology is still being developed, it is unclear what the long-term
effect of such systems will be. Previous research into automated highway systems
suggests some questions that are also applicable to the application of electronic guidance
of urban buses. Of those questions put forward by Levitan ( 1998) we suggest the
following two are relevant to electronic guidance:
a) What effects will automated travel have on manual driving?
b) What role can the driver be expected to play when a failure occurs?
16
1.3.4 Driver Vehicle Interface ( DVI) Design Issues
Electronic guidance systems should be designed to be as easy to use as cruise control in a
car. It is expected that the guidance function will be engaged with simple actions and
that the driver will be given clear confirmation regarding the engagement/ disengagement
of electronic guidance and system status.
The driver will interact with the electronic guidance system through a Driver- Vehicle
Interface ( DVI), which will provide the necessary information and the means to make an
informed decision regarding transition to or from electronic guidance at any given time.
The DVI may include a visual display ( which could be a simple array of LED’s, an LCD
screen, or a Heads Up Display), auditory communication, tactile feedback ( such as a
vibrating seat), and haptic feedback ( such as brake pulsing), as well as a set of simple
controls. The design of the system should promote good mode awareness, which means
that individual switches and displays should not have different meanings at different
times such that they could cause driver confusion. Furthermore, the design of the system
should assist the driver in recognizing, diagnosing, and recovering from errors. It is
important to note that the DVI should be designed to support the goal of reducing driver
workload in challenging driving conditions
Most, if not all, deployments of electronic guidance, will involve route sections where the
infrastructure support for guidance is available and sections where it is not. For this
reason, a smooth transition from manual to assisted driving and back is essential. The
transfer of control could potentially be achieved in two ways. The first is driver initiated,
requiring some action from the driver to engage or disengage the auto steering mode.
The second method is a system- initiated transfer to/ from auto mode. Even when the
transition from auto mode to manual is automatic, say at the end of a narrow tunnel, the
driver must take some sort of action to let the system know that he is back in control. If
the driver does not take the required action to switch back to manual steering, the vehicle
may come to a complete stop. An example of a driver initiated auto- steering handover
would be as follows:
a) As the bus approaches narrow sections of roadway the system indicates to the
driver that they are in an auto- steering enabled area.
b) The driver decides to transition to auto steering and acts to engage the auto-steering
mode.
c) While in auto steering mode the driver controls speed and monitors the forward
and side views.
d) Prior to the end of the narrow roadway section, the system indicates to the driver
that the bus is approaching the end of the auto- steering enabled area.
e) The driver takes action and transitions to normal steering. If the driver does not
take the required action to switch back to manual steering the vehicle comes to a
complete stop.
In case of an emergency, the driver will have the ability to exit or override the system in a
number of quick and simple ways such as pushing down on a button, applying a large
17
torque to the steering wheel, or applying hard braking. Redundancy and system fault
detectors need to be built into the system so that if a system fault occurs, the system can
switch to a back- up method of control and indicate to the driver that the system requires
maintenance.
A number of human factors recommendations for automated systems come from research
on automated highway system ( AHS). The guidelines below are taken from FHWA- RD-
97- 125. We have taken only the guidelines that we feel would apply to transit bus
electronic guidance implementations.
• If the driver must accurately position the vehicle to be able to transition, then a
guideline should be painted on the roadway or an in- vehicle display should
graphically depict the vehicle position relative to a reference point.
• Establish mechanical restrictions to guard against control transfer until the driver is
prepared to initiate it.
• Provide attention- getting displays ( e. g., rumble strips) for areas where an automated-enabled
segment of roadway ends.
• If the driver disregards an alarm the vehicle should come to a controlled stop.
It is important to acknowledge that implementing electronic guidance to urban bus
driving will change both the driving task and the operational environment. Such changes
in any system can have both positive and negative impacts. It is therefore important to
clearly define what the system is prior to implementation and to determine what the
expected changes might be so that any potential negative impacts can be designed out.
Urban bus operation can be thought of as a system comprised of five main entities: the
bus, other traffic and operating conditions, the operator, the passengers, and the transit
agency. Each of these entities has the ability to influence all the other entities. For
purposes of this section we have looked specifically at the bus operator’s task, though it
is important to recognize that the other entities have the ability to impact the operator’s
task.
1.4 Terminology Definitions
Electronic guidance and lane assist have been interchangeably used by different people
for different applications. In this report, “ electronic guidance” refers to technologies that
provide automated steering or driver assist functions enabling a vehicle to follow a
certain predetermined trajectory under automatic control. The term “ lane assist” refers to
the application of electronic guidance technologies to allow a transit bus to maintain a
transit vehicle in a designated lane or a desired trajectory while “ precision docking”
refers to application of electronic guidance technologies to deliver accurate, reliable and
repeatable maneuvers that allow safe, convenient, and expedient boarding and alighting
operations at bus stations constructed in a train- platform manner. Electronic guidance
can be combined with longitudinal control. It is noted that for certain applications, it is
advantageous to integrate both the lateral and longitudinal functions for performance
requirements. For instance, precision docking that demands accurate positioning of
18
vehicles laterally and longitudinally may be difficult to accomplish with drivers
controlling the stopping maneuvers.
1.5 Report Organization
This report addresses the needs that can be served by lane assist systems and the
requirements that these systems will need to meet in order to be found beneficial. It is
intended to be complementary to the report that was previously produced by the
University of Minnesota for FTA and Metro Transit ( Donath, et. al., 2003), adding further
information about safety issues and incorporating the findings from close interactions
with stakeholders at key transit properties who could become early adopters of lane assist
systems.
Chapter 2 reports on the findings of the case studies that were performed with AC
Transit, LACMTA, the Lane Transit District and San Diego Transit to identify their
needs and requirements for lane assist systems.
Section 3 provides a benefit and cost analyses showing the needs of the transit lane assist
and precision docking systems and an economic analysis of a subset of the economic
benefits that can be gained from lane assist systems ( dwell time savings at bus stops for
precision docking and reductions in busway width for automatic steering). Deployment
issues are also discussed in this chapter.
Chapter 4 provides a preliminary description of performance requirements and technical
specifications for transit lane assist and precision docking systems.
Chapter 5 is the summary.
Additional information is provided in Appendices as follows:
Appendix A: Effects of Tight Turning Radii on Needed Lane Width
Appendix B: Questions for Lane Assist Requirements Workshop
19
2.0 CASE STUDIES
In order to study the needs and define the requirement specifications for transit lane assist
and precision docking systems, the project team studied the potential application of lane
assist technologies for, and conducted a series of workshops at, four partner transit
properties, including:
• Workshop with Lane Transit District ( LTD) in Eugene, OR on July 29, 2003
• Workshop with San Diego Transit and San Diego Association of Governments
( SANDAG) in San Diego, CA on September 10, 2003
• Alameda- Contra Costa Transit District ( AC Transit) Workshop at PATH facility
in Richmond, CA on February 12, 2004
• Workshop with Lane Transit District ( LTD) drivers in Eugene, OR on March 31,
2004
• Workshop with Los Angeles County Metropolitan Transportation Authority
( LACMTA) in Los Angeles on September 3, 2004
• Combined Workshop at University of California Richmond Field Station on July
29, 2004
The purposes of these workshops were to acquire first hand information from transit
operators and drivers, to derive a common set of needs, benefits and costs, and to
translate these needs and desires into system requirements and specifications.
The case studies and workshops differed considerably across these transit properties
because of the different maturity of development of the BRT concepts within these
properties. Specifically, Lane Transit has already progressed further in its own
development of its BRT system and in its consideration of lane assist technologies than
the others, so this case study had significantly more depth than the others. In contrast,
AC Transit is in the midst of planning for its new BRT service but has only begun
consideration of the opportunities provided by lane assist. LACMTA already has a
number of BRT lines in operation and is considering whether lane assist and precision
docking technologies can be beneficial to improve the operation. San Diego Transit, on
the other hand, is just starting to think about where it could apply BRT service, so it had
not begun considering lane assist possibilities prior to the project workshop.
The formats of the workshops with the four transit properties were similar, beginning
with an overview presentation of lane assist and precision docking systems and following
with discussions of various issues related to the needs, requirements and deployment of
these technologies. The discussions were led by PATH researchers. An exception was
the AC Transit workshop, which also provided participants with the opportunity to drive
an automated bus developed by PATH. One additional workshop was organized by LTD
to focus on obtaining inputs from drivers.
As the culmination of the project, representatives of the participating transit properties
came together in a one- day workshop at the University of California Richmond Field
20
Station on July 29, 2004, followed by a teleconference on August 3, 2004, to address
issues that could not be covered within the workshop schedule. These were opportunities
to provide comments on the draft project report, with particular emphasis on the
definition of requirements for lane assist systems. More importantly, this was the
opportunity to gain cross- fertilization of ideas from transit properties that have
approached lane assist systems from somewhat different perspectives, and to see whether
a common set of requirements could address all their needs. The workshop was attended
by representatives of AC Transit, Lane Transit District and San Diego Transit, and they
were joined on the teleconference by a representative of LACMTA. Those who attended
the workshop had the opportunity to drive a test bus equipped with precision docking
capability, so they could experience it in action from the driver’s perspective.
The inputs obtained from the representatives of the four transit agencies through six
workshops were organized into key issues including applications, benefits and costs,
design considerations, requirements, and deployment issues for lane assist and precision
docking systems.
2.1 Potential Applications
Among the four partner transit agencies, Lane Transit and AC Transit have already
considered how to incorporate lane assist and precision docking systems in the designs of
their current BRT systems. The project team therefore conducted more detailed case
studies on the LTD and AC Transit BRT corridors, while potential applications of lane
assist and precision docking systems for LACMTA and SANDAG were discussed in
their workshops.
2.1.1 LTD Case Study Example
As a result of a major investment study conducted between 1992 and 1999, Lane Transit
District ( LTD), which serves the towns of Eugene and Springfield, Oregon, is in the
process of implementing a Bus Rapid Transit ( BRT) system. It is based on light- rail
transit operating principles, but uses buses in service that is integrated with key
components of the existing automobile transportation infrastructure, such as roads, rights-of-
way, intersections, and traffic signals. This system is more affordable and flexible
than light rail and allows for incremental construction and implementation which can be
easily tailored to meet the specific transportation needs and opportunities within
individual neighborhoods and transportation corridors. BRT offers many advantages
compared to regular bus service for LTD, including service frequency, increased
capacity, and speed.
One of the key features of the BRT system, reduced dwell times at stops, will be achieved
primarily by use of off- board fare collection. It is hoped that precision docking can be
added in the future. By creating a bus platform at the same height as the floor of the bus
and eliminating a gap between the platform and the bus, all passengers can board easily
and quickly.
21
If LTD’s BRT system is to achieve its service goals, some form of electronic guidance
system will be needed to allow it to operate in narrow right- of- ways and to precision
dock. Although the BRT buses have been ordered, no guidance system has been chosen.
Proposed station spacing is approximately every half- mile, resulting in ten stations, eight
of which will be new facilities. Stations will be similar to light rail stations, with a high
level of passenger amenities, including benches, shelters, ticket machines and passenger
information.
BRT in Eugene and Springfield will be developed incrementally in order to adjust to
community needs and as resources become available. The pilot corridor will be the route
from the Eugene Station at Willamette Street and West 11th Avenue in Eugene, to
Franklin Boulevard, past the University of Oregon, and along South A Street in
Springfield to the new Springfield Station. This corridor will introduce the service to the
community, offering bus riders and motorists first- hand experience with its benefits.
Future routes that will connect to the pilot corridor are currently under consideration for
both Eugene and Springfield.
Consistent with the approved design, there will be exclusive transit lanes from the
Eugene Station to Walnut Street. From Walnut Street to McVay Highway, the BRT
system will travel in mixed traffic. Glenwood is expected to redevelop over the next 10-
20 years, and it was felt that until the redevelopment occurs, it does not make sense to
spend limited funds on roadways that are likely to change. However, once across the
bridge into Springfield, BRT will travel in exclusive transit lanes between the bridge and
the new Springfield Station at Pioneer Parkway East and South " A" Street.
Eight BRT stops between the Eugene Station and the Springfield Station are constructed.
All traffic signals along the route will be designed to give priority to the BRT vehicles.
The second Eugene BRT corridor will be Coburg Road, originating at the Eugene station
and running over the Ferry Street Bridge on Coburg Road to Crescent Avenue.
Alternative street routing will be examined as part of the design process. Connections to
the Pioneer Parkway corridor, being developed in Springfield, will also be pursued.
The City of Springfield and LTD have selected the Pioneer Parkway corridor to be the
next BRT corridor developed in Springfield. The Gateway area has seen extensive
employment growth and increases in traffic congestion. More frequent and reliable
transit service is necessary to serve this growing area. The City of Springfield is planning
an extension of Pioneer Parkway, north of Harlow Road, to serve the new Sacred Heart
Hospital and Riverbend development area.
22
Figure 2.1 LTD Proposed BRT Corridors
LTD has identified the New Flyer BRT vehicle as having the desired vehicle attributes,
such as alternative power, clean, quiet, low- floor, tram- like in appearance, for service in
the Eugene/ Springfield area. The ability of a vehicle with lane assist capability to
maintain a predefined path throughout the corridor would result in less right of way
needs, increased operational efficiency, and easier passenger access/ egress.
LTD believes that vehicle guidance is the next major development in transit.
Specifically, vehicle guidance will solve the following specific problem areas:
• Limited Right of Way: The vehicle is able to follow a pre- programmed route;
thereby minimizing lane width requirements and increasing operating speeds while
reducing the amount of right of way needed.
• Dwell times: The guidance system will allow the vehicle to stop very close to a
passenger platform. When combined with a passenger platform that is raised to the
height of the floor of the vehicle, boarding by all passengers, including people in
wheelchairs, is accomplished easily and quickly and without the need for special lifts
or ramps.
23
• Image: The guidance system will allow the vehicle to operate in a similar manner to
a rail- guided vehicle. This image, combined with the other BRT features, will attract
a new market of riders who have traditionally been reluctant to use conventional bus
service.
• Maintenance costs: Smooth acceleration and deceleration will enhance passenger
comfort and also emulate rail operations. Over- braking and acceleration of vehicles
will be reduced, thereby decreasing vehicle maintenance.
( Inputs from LTD workshop, July 29, 2003):
LTD foresees deploying electronic guidance at stations, stops, intersections, and along
Franklin Boulevard, which may require contra- flow ( narrower) lanes. On this route, high
speeds are not required ( need only 30- 35 mph). If the system handles both steering and
speed control, it would achieve its most efficient operation, but then it may be difficult to
keep the driver engaged in the driving process. Operation involving electronic guidance
will be primarily on the median side of traffic in exclusive rights- of- way and, where
possible, segregated.
2.1.2 AC Transit Case Study Example
AC Transit ( the District) operates over 100 local and Transbay bus routes. Six of these
routes carry over 15,000 daily riders. The District’s Bus Rapid Transit project is planned
for their heaviest route, which today carries 30,000 daily riders. The BRT corridor would
serve three downtowns ( Oakland, Berkeley and San Leandro); three regional medical
facilities ( Alta Bates, Summit Medical Center and San Leandro Hospital); the 30,000
student University of California, Berkeley campus, Vista Community College and the
Bayfair Mall. The residential density in the corridor varies between 11,000 and 23,000
persons per square mile. The corridor serves 1/ 3 of the residents and half the jobs in
Oakland. The project currently has $ 100 million in local and regional funds.
The entire 18- mile corridor would use existing arterial streets and would convert two 12-
foot traffic lanes into dedicated bus lanes. Much of the route is 75 feet wide ( curb to
curb). In some cases, the alignment is also the most direct route for bicyclists and is
designated as a future Class II bikeway in the Oakland Bicycle Master Plan. The
conflicting demand for road space has motivated the District to seriously consider lane
assist solutions that could permit narrower bus lanes.
AC Transit would like to explore opportunities to implement the magnetic guidance
technology along AC Transit’s BRT corridor. AC Transit sees that lane assist technology
can provide the following benefits:
• Accommodation of bike lanes for the entire route
• Accommodation of additional traffic lanes at congested locations
• Reduced infrastructure cost ( approximately 12% of busway costs)
24
• Increased safety
• Preservation of on- street parking in local commercial districts
• Reduced dwell times and enhanced ease of boarding and alighting at BRT
stations. Since the new buses will load wheelchairs at the center door, positioning
the bus becomes both more important and more difficult. Precision docking could
be critical here.
In addition, AC Transit views longitudinal control as a potentially important component
of its Bus Rapid Transit program and key adjunct to the lane assist
technology. Longitudinal control will enable smooth accelerating and braking that will
arguably permit buses to operate with the comfort and safety of rail. From the
passenger’s perspective, longitudinal control would minimize the harsh movement of the
bus and reduce on- board passenger falls. Longitudinal control would also reduce wear
and tear on the bus’s drive train and brakes, reducing maintenance costs and fuel
consumption.
( Inputs from AC Transit workshop, February 12, 2004):
AC Transit sees that lane assist and precision docking systems can enhance its new BRT
corridor along Telegraph/ International/ E. 14th St. These technologies may also be
implemented along the existing San Pablo RAPID BRT line.
The subject of which technology would be best was brought up and discussed briefly.
Most of the discussion dealt with magnet systems, which seemed the most flexible for the
applications envisioned. One person thought that a vision system, with its highly visible
painted line, would be better in that people could see easily see the busway. It was
pointed out that the same effect could be achieved with magnets by simply painting a line
connecting them. Another possibility would be to use a different type of pavement or to
paint the entire busway surface a different color.
2.1.3 SANDAG’s Considerations in Lane Assist and Precision Docking
Systems
( Inputs from Sandag workshop, September 10, 2003):
The evaluation of lane assist systems is most closely linked and can be appropriately
considered in conjunction with the “ Transit First Showcase” project that is administered
under SANDAG. Particular features of the targeted showcase project are in line with the
characteristics of transit electronic guidance systems, including:
( 1) Exclusive right- of- way to bypass congested areas, and use of shared lanes where
streets are operating smoothly. The portion of the route that is designed for
exclusive transit lanes is El Cajon Blvd. from Park Blvd. to 43rd Street.
( 2) Stations will have uniform design features, with accessibility to passengers with
disability, in compliance with the Americans with Disabilities Act.
25
( 3) Offering ride quality of rail and operating qualities of buses, and ultimately
increasing ridership considerably.
Lane assist systems could enhance the service of transit buses and become an integral
part of the developments for the following aspects:
• Train- like ride quality
• Service quality and reliability
• Transit station design and ADA compliance
• Construction cost savings in exclusive transit lanes.
At the time of this workshop, San Diego area stakeholders had not yet focused enough
attention on the lane assist technologies to be able to offer specific feedback regarding
their needs, priorities or concerns related to lane assist, limiting the value of the current
case study.
2.1.4 LACMTA’s Considerations in Lane Assist and Precision Docking
Systems
( Inputs from LACMTA workshop, September 3, 2004):
During the discussions several areas within the MTA’s current route system were
discussed as potential locations where a lane assist and/ or docking system could be used:
• A strip of 13 miles on Wilshire Boulevard. This is the corridor where most of the bus
accidents happen ( due to the higher number of miles driven in this corridor). A recent
effort using the Sheriffs department to increase traffic enforcement saw a drop from
11% of all accidents to 9% of all accidents. Typically the Rapid buses have 30%
fewer accidents than local buses ( on a per mile driven basis)
• The Orange Line, which is 14 miles long, 13 miles of which are on an exclusive lane
constructed in former rail right- of- way. The other mile is on an arterial in mixed
traffic using transit signal priority to expedite travel time. There are 22 new NABI
articulated buses providing the service on this corridor. There are 13 stations ( with the
eastern most station being the North Hollywood MetroRail Red Line station). There
is a parallel bike path, as well as new parking facilities at five of the stations, but
costs associated with these items should be the same with or without lane guidance.
Stations are approximately one mile apart and buses will have 5- 10 minute headways
during the peak periods. Since the curb on this line will be at the same height as the
bus floor this could be used as a demonstration project for precision docking.
• San Fernando Valley demonstration line – there is a section going under the I- 405
freeway where there are tight tolerances between columns, so precision docking could
be useful.
• Line 710 has a new concrete roadway. There was discussion about the height of the
footpaths here, with no consensus, but it seemed that the difference between the curb
and the floor height of the Nabi articulated buses’ front door was not that large.
26
There was consensus that the time needed to deploy wheelchair ramps was not great
and that level boarding was not necessarily going to be a huge advantage or need.
• Sections where narrow lanes already exist. Three sections of the Wilshire Metro
Rapid Corridor, including a segment of roadway in Santa Monica where tree- lined
medians have recently been constructed, in Beverly Hills, and east of the “ Miracle
Mile” between La Brea and Western Avenues in Los Angeles, already have lanes as
narrow as 10- 10.5 feet wide.
2.2 Benefits and Costs
2.2.1 Benefits
( Inputs from LTD workshop July 29, 2003):
In the LTD service area there are quite a few people with disabilities. Precision docking
would allow much faster boarding. Additionally, in an informal poll, 15 of 17 attendees
at one of the workshops preferred electronic guidance to mechanical lifts for wheelchair
access, with frequent lift maintenance cited as one of the principal reasons.
If BRT and electronic guidance are able to reduce travel times, increase reliability, and
eliminate the second- class stigma of bus travel, ridership should increase significantly,
thereby reducing congestion, pollution, and commute times, and increasing farebox
returns. Lane assist and precision docking also have the potential to reduce collisions as
well as passenger injuries.
Workshop participants discussed the fact that, since subsidized public transit is the norm,
what sort of return on investment should be required in order to justify the expense of
electronic guidance? The participants felt that while forecasting can provide at least an
estimate of increased revenues due to increased ridership as well as savings from
increased reliability and safety, there is no easy way to quantify such intangibles as
quality of life gains from reduced congestion and commute times, a cleaner environment,
and reduced stress.
Workshop participants felt that LTD must get “ reasonable payback” in about 10 years,
with initial costs ( which it was believed would be paid, for the most part, with federal
funds) and ongoing operating costs as the most important to consider. It was further
hoped that LTD could break even on traditional operational costs, including those
involving infrastructure, with delay reductions and increased reliability providing a
portion of the cost reductions. Another potential source of operating cost reduction
would be increases in the lifecycle of other bus systems. Brakes, for example, may last
longer under electronic guidance.
( Inputs from AC Transit workshop, February 12, 2004):
A commonly discussed benefit in this and other workshops is increased trip time
reliability and the saving of time. These could be achieved by reducing dwell times
27
through faster boarding of wheelchairs and other riders with mobility problems, queue
jumping, and signal priority. Time saved through faster boarding could be significant,
since census information along the corridor ( in an area around the length of the corridor
with a width equal to the distance people are normally willing to walk to bus stops)
shows that the percentage of disabled persons is twice the average of the Bay Area as a
whole. An ancillary benefit to faster boarding is that it would help resolve a stressful
issue for drivers who must deal with passengers whose patience is often strained by the
time it takes to load and secure wheelchair passengers.
Funding is always a problem for public transit agencies and AC Transit is no exception.
To sell the system, benefits would have to be significant. Bundling electronic guidance
with other systems such as collision warning and speed control would be seen as a much
more attractive package.
While time and reliability benefits are important in attracting new riders, for transit
agencies an important question when considering new technology is not simply how to
decrease dwell times at individual stops but whether the total time saved will allow them
to decrease the number of buses needed to service the route. This is where the real cost
savings would come into play.
Passenger falls when the bus stops are a problem for AC Transit. The busy ( trunk) lines
have a high probability of standees, which makes the problem worse. Any technology
that could provide smooth, predictable acceleration and deceleration would have the
potential to reduce the number of claims against the agency, which could produce
significant cost savings. Certainly, precision docking would help in this area since
station stops would be consistent and predictable. Additionally, a smoother, “ rail- like”
ride might make it possible to do away with the requirement to secure wheelchair
passengers. However this would require that guidance be used on the entire route,
something which is not feasible in the foreseeable future.
Two additional potential benefits were also mentioned. The first was the possibility that
precision docking would lower some of the ADA mandated door- to- door paratransit
service costs by allowing delivery of passengers to fixed- route buses rather than door- to-door.
Second, guidance might allow lane width to be decreased from 12 to 10 feet,
leaving more space for other purposes such as parking or bicycle lanes. It may be
difficult to monetize this when calculating benefits, though.
The ultimate measure of success of a transit system is whether or not people are using it.
The measure of success for a transit system change is therefore its effect on ridership.
This can be difficult to measure, since it is often impossible to isolate the specific effect
of the change in question. The implementation of vanpools, for example, was
accompanied by fare adjustments, service changes, and balky fareboxes. On a positive
note in relation to BRT, a before- and- after study is being conducted on the San Pablo
RAPID corridor and, thus far, it appears that there has been a 12 % increase in ridership
since the start of the new BRT service there.
28
( LTD driver workshop, March 31, 2004):
Drivers generally are in favor of lane assist and precision docking systems and believe
that these systems offer the following benefits:
• Enables fast at- grade boarding ( reducing dwell time)
o Access for riders made easy and more efficient
o Enhances boarding for wheelchair passengers
• Enables buses to operate within narrow lanes that facilitate dedicated BRT
deployment without increasing operator stress
• BRT deployment may lead to an increase in ridership, if riders experience faster
transit times.
Two drivers commented that they felt it was the “ way of the future,” one driver said he
“ did not want to see a BRT application implemented without it”, and the drivers also felt
that if the system worked properly it could reduce their stress level. The drivers felt that
the most beneficial aspect of the system for their application would be the docking.
( Inputs from combined workshop, July 29, 2004):
There was general consensus that the benefits of Lane Assist need to be stated very
clearly. Most important to LACMTA are safety, reliability, and consistency. Reducing
dwell time is not at all a selling point for them. They cautioned that the theoretical dwell
time savings can get watered down considerably when it is filtered through the entire
staffing and rostering process. It is important that all steps in the operations process be
accounted for so that a more realistic assessment of benefits can be made.
Safety and liability were major topics of discussion. One of the risk management
participants strongly disagreed that the introduction of a lane assist system would see a
large reduction in their accident costs, which were estimated at $ 40 million in 2003. Of
this amount, approximately half was the result of severe injuries to pedestrians who were
hit by a bus as it made a right turn. These type of accidents cost somewhere in the range
of $ 1.5 to 4.5 million each.
While pedestrian accidents would not be helped by lane guidance, this still leaves $ 20
million worth that could be. Paradoxically, it was felt by participants that there could be
an increase in liability for prior accidents when safety improvements are added ( it is seen
as your company acknowledging that there was a hazard that should have been addressed
sooner). This actually occurred when the agency introduced a 4- quadrant gate at several
intersections. It was noted, however, that LACMTA would not reject a new technology
system simply because of the possibility of liability for prior accidents.
Bus versus a stationary object is by far the most frequent crash type, although these
crashes usually have low direct costs as they occur in crowded streets at low speeds.
However there are many hidden or indirect costs ( administrative, loss of bus in service,
involvement of 3 to 4 transit operations supervisors, union involvement) for these
29
accidents which contribute significantly to the actual total. Thus the little, every day stuff
of 5- 6 accidents/ day ( a few thousand annually) quickly add up.
Because there exists the potential for a lane guidance system malfunction to cause a
crash, one of the risk management participants noted that in order for the LACMTA to do
a demonstration line they would require that the manufacturer or the funding agency have
$ 50 million of insurance to cover the risk. At present when MTA purchases any
equipment for a bus, it has to come with $ 20 million of insurance. The extra $ 30 million
would be required for the perceived extra risk of the new technology. This issue could
overwhelm a startup company trying to get into this business. We will need case law on
this and should closely monitor the Las Vegas BRT.
The biggest economic benefit a large transit property would see from time saving
associated with exclusive busway operations or precision docking would be if it became
large enough to eliminate a bus from a major route without diminishing quality of service
( estimated at $ 250 K per year for 16 hours of daily operation by AC Transit). However,
this was of less value to Lane Transit because their services need to operate in multiples
of 60 minutes to maintain a regular schedule and they could not save a bus.
It is important to have a way to factor the time savings from precision docking into the
transit agency estimates of hourly operating costs so that it can be compared to other
alternatives. The AC Transit General Manager is excited about potential time savings
from docking, but needs a better way of quantifying that. LACMTA has recently
deployed a 14- mile BRT facility to standards comparable to an LRT line, with full- scale
stations, in the San Fernando Valley. It could be considered as a test site to compare
dwell times with and without precision docking.
2.2.2 Costs
( Inputs from LTD workshop July 29, 2003):
It is anticipated that there will be $ 1.6 M that won’t be spent on vehicles. At this time, it
is not known how the decision will be made regarding spending the money on electronic
guidance. A final question considered was whether or not it would make sense to equip
the entire fleet with guidance.
( Inputs from combined workshop, July 29, 2004):
The participants felt there was a need to work with transit planners to determine cost
savings. It was emphasized that the only way for time savings brought about by new
technology to pay off is if the savings are large enough to allow a bus to be taken out of
service while maintaining the same level of service.
First, the following costs, as they relate to any new technology, must be estimated:
• Capital
• Operating
30
• Maintenance
• Installation
• Training
• Upgradeability/ lifecycle
Then do a calculation for, as an example, 100 buses on five miles of Wilshire Boulevard.
What would the saving be for one year, two years, and beyond ( it would be expected to
lose money for the first few years)? What’s it going to save over the life of the system
( it would need to last more than 10 years)? How many miles would you need to
implement to be able to take a bus out of the system?
It was stressed that what is really needed is to have more capacity for the same operating
costs, which would need to involve scheduling system experts in the analysis. Since
ridership is largely transit dependent for buses ( and rail), lane assist technology, by itself,
may or may not attract more people since it will be invisible to most people. What is
needed is to increase ridership without increasing costs. Putting more people on an
already crowded system could actually cost money since another bus would eventually
need to be put into revenue service.
Costs were also recognized to be in a different category from system requirements, but
there are clearly going to be trade- offs that determine what costs will be acceptable or
unacceptable for deploying lane assist systems. The transit properties will need to be
able to judge the costs of the systems against the benefits ( financial and other) that they
will gain from use of the systems. The discussion gravitated more toward the benefit side
than the cost side.
Cost estimates need to be done carefully to separate the cost increments that are
specifically associated with use of the lane assist system. For example, in the case of
precision docking at stops or stations, the entire cost of the stop or station should not be
associated with the docking function, but only the cost elements that are different because
of precision docking.
2.3 Requirements
2.3.1 Performance
( Inputs from LTD workshop July 29, 2003):
With specific applications in mind, what kind of performance can LTD expect to realize
from electronic guidance? Additionally, should these expectations be judged by current
bus or light rail standards?
As a starting point, the system should be easy to operate within the full range of
applications. Next, ride comfort, which is best defined as smoothness of operation, is
very important especially since LTD anticipates fewer available seats and more standees
( some with bikes) on board than on standard buses. This being the case, smooth
31
acceleration, deceleration, and lane tracking are critical. At a minimum, performance
under electronic guidance during these phases should be better than in the manual mode.
Tracking accuracy is important under two application scenarios: operations in narrow
lanes and precision docking. Since Franklin is an Oregon state route, ODOT is very
interested in the BRT lane width along this route and the ability of the guidance system to
stay within it. Several workshop participants wanted to know if an S curve or a series of
curves would require a greater lane width, not only to compensate for rear wheel tracking
but also electronic guidance sloppiness.
For docking, ADA requires a distance no greater than two inches horizontally and 5/ 8
inch vertically between the bus entrance and the station platform. This presents a special
problem for articulated buses since without a long straight approach, the rear of the bus
would not be in alignment with the front. This would necessitate a specific platform
shape, with the aft part angling out so as to align with the rear bus section.
( Inputs from AC Transit workshop, February 12, 2004):
For a number of reasons, some more perceived than real, people like rail better than
buses. One reason is that trains provide a smooth ride, with few, if any, unexpected
movements. To be more “ rail- like,” workshop participants felt that the system should be
smooth enough that it would “ not spill my drink.” Passengers should not have to contend
with any unexpected movement. The smoothest ride would be provided by a system that
had longitudinal as well as lateral guidance. The demonstration ride at the beginning of
the workshop seemed to fit this need.
The following categories of performance attributes were offered for consideration:
2.3.1.1 Lane- tracking accuracy
( Inputs from LACMTA workshop, September 3, 2004):
Opinion was split about how important it was to have minimal lateral deviation ( less than
5 cm) at high speeds. It was agreed that a high accuracy system should be available, as it
is critical on the arterial routes.
( Inputs from combined workshop, July 29, 2004):
Increased lane- tracking accuracy makes it possible to operate the bus in a narrower lane.
Both Lane Transit and AC Transit thought that the narrower the feasible lane width, the
better in general, but they