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 III: Interface Requirements
UCB- ITS- PRR- 2007- 23
California PATH Research Report
Fanping Bu, Wei- Bin Zhang, Susan Dickey,
Steven E. Shladover, Han- Shue Tan
CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS
Lane Assist Systems for Bus Rapid
Transit, Volume III: Interface
Requirements
Fanping Bu, Wei- Bin Zhang, Susan Dickey,
Steven E. Shladover, Han- Shue Tan
Prepared for:
California Department of Transportation
Federal Highway Administration &
Federal Transit Administration
Prepared by:
California PATH Program, University of California at Berkeley
Lane Transit District
AC Transit
Final Report for RTA 65A0160
ii
iii
Acknowledgements
This work was performed by the California PATH Program at the University of California at
Berkeley in cooperation with the United State Department of Transportation Federal Transit
Administration and State of California Business, Transportation and Housing Agency,
Department of Transportation ( Caltrans) under Federal ID # CA- 26- 7034- 00 through RTA
65A0160. 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 and the United State government.
This report presents the results of a research effort undertaken by the University of California
PATH Program, Lane Transit District and AC Transit, sponsored by the Federal Transit
Administration under, with cost share from the California Department of Transportation. The
United States Government and California Department of Transportation assume no liability for
its contents or use thereof. The United States Government and California Department of
Transportation does not endorse products of manufacturers. Trade or manufacturers’ names
appear herein solely because they are considered essential to the objective of this report.
The direction of Brian Cronin, Sébastien Renaud and Venkat Pindiprolu of the Federal Transit
Administration and Yehuda Gross of the Federal Highway Administration, ITS Joint Program
Office, is gratefully acknowledged. Special thanks are also due to the California Department of
Transportation ( Caltrans) for providing additional funding and contractual assistance. We would
like to thank Sonja Sun and Don Dean for their assistance and support. The assistance and
feedback of Mathew Hardy of Mitretek has also been beneficial to the effort of this research and
evaluation program.
Also, this work would not have been possible without the cooperation of the transit agencies.
Specifically, Graham Carey and Stefano Viggiano of Lane Transit District ( LTD), and Jim
Cunradi of Alameda Contra Costa County Transit ( AC Transit) have contributed greatly during
the technical visit to Europe and to this technical report.
The authors would like to express their sincere appreciation to many people in the following
companies who supplied information about their buses:
• Gillig Corporation
• New Flyer Industries
• VanHool
iv
v
Abstract
Vehicle Assist and Automation ( VAA) systems enable lane assist, precision docking and
longitudinal control of transit vehicles. They offer the opportunities of providing high quality
transit service within reduced lane widths. Transit vehicles in North America are mostly
manufactured based on individual transit agencies’ customized requirements. The interfaces
between VAA components and the mechanical, electrical and electronic systems on the existing
transit vehicle, if not defined properly, can be an impediment to large scale deployment of VAA
technologies. This report summarizes a research effort in specifying the VAA interface
requirements, with a goal to facilitate progress toward the development and deployment of VAA
systems on transit vehicles in the U. S., so that transit agencies and their passengers can start to
experience the benefits these systems can provide.
Keywords: Vehicle Highway Automation, Lane assist, electronic guidance, Bus Rapid Transit
vi
vii
Executive Summary
The interface requirements defined here are intended to facilitate progress toward the
development and deployment of Vehicle Assist and Automation ( VAA) systems on transit buses
in the U. S., so that transit agencies and their passengers can start to experience the benefits these
systems can provide. VAA systems offer the opportunity of providing high quality transit service
within reduced lane widths. VAA includes four functions allowing buses to perform precision
docking at bus stations, vehicle guidance or automatic steering on the running way between
stations, automatic platooning of buses at close separations and fully automated vehicle
operations. The precision docking functions facilitate passenger boarding and alighting at
stations, while vehicle lateral guidance could support reduced lane width, allowing the bus to
operate in a designated lane that is only slightly wider than the bus itself without increasing
driver workload. It could be implemented in 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. Transit operators are very interested in VAA
in order to deliver rail- like service, an attractive feature to riders, at a fraction of the cost of rail.
The primary technological barrier to VAA deployment is the fact that many of these VAA
products are tied to a specific, specialized and costly vehicle and cannot be easily retrofitted onto
the existing buses produced by North American bus manufacturers. Whether a U. S.- based or
imported VAA system is considered, the prerequisite is that the VAA system must be able to
interface with existing bus subsystems.
Transit vehicles in North America are mostly manufactured based on individual transit agencies’
customized requirements. The interfaces between VAA components and the mechanical,
electrical and electronic systems on the existing bus, if not defined properly, can be an
impediment to large scale deployment of VAA technologies. Therefore, there is a great need to
understand how VAA systems, based on any practical technology, can interface with transit
vehicles and infrastructure. A standard set of interface requirements will be needed to allow the
suppliers to develop VAA technologies with common interfaces and to allow bus manufacturers
to retrofit the VAA technologies of transit agencies’ choice to different buses without excessive
custom design work or modifications to the existing products. These interface requirements are
critical to both vehicle manufacturers and suppliers to achieve compatibility, ease of safety
verification/ certification and to lower cost and reduce deployment time. To address these needs,
the U. S. Department of Transportation, through the Federal Transit Administration ( FTA) and
the ITS Joint Program Office ( ITS- JPO) sponsored the project reported herein to study the VAA
interface requirements.
In order to clearly define and identify the interfaces between VAA subsystems/ elements and bus
subsystems/ elements, a modular system architecture was established to analyze VAA system
functional blocks and information flows. This modular architecture defines the nature of the
interface between the VAA system and the bus and bounds the physical interface to a small
possible set, therefore is essential for the development of the interface requirements. Based on
the modular VAA architecture, interface requirements were developed. The interfaces between
viii
ix
the VAA system and an existing transit bus were classified into three major categories:
mechanical, data communication and power supply. Functional units of the VAA system ( e. g.,
lane positioning sensing, vehicle state sensing, steering actuator, brake actuator and propulsion
actuator) were analyzed to determine their interface requirements with existing bus subsystems.
Special attention was devoted to the data communication, which is the backbone of the proposed
VAA system architecture. The interface requirements between buses equipped with VAA
systems and infrastructure were also studied in terms of running way width, sensed infrastructure
references, boarding platform and vehicle exterior geometry.
Finally, experiments were conducted on an advanced BRT vehicle previously developed by
PATH under Caltrans sponsorship, to validate the interface requirements. The experiments were
focused on the requirements for data communication, since these cannot be verified by simple
inspection of designs or drawings. The SAE J1939 protocol that is popular in the heavy- duty
vehicle industry was used for the data communication of the tested system. Aspects of the data
communication such as timing and data length were studied.
x
xi
Table Of Contents
Acknowledgements ................................................................................................................... iii
Abstract ............................................................................................................................... ...... v
Executive Summary ................................................................................................................. vii
List of Abbreviations................................................................................................................ xxi
1.0 Introduction ................................................................................................................. 1
1.1 NEEDS FOR VAA INTERFACE REQUIREMENTS ................................................................................................. 3
1.2 INTERFACE REQUIREMENTS VS. SYSTEM REQUIREMENTS AND DESIGNS ..................................................... 7
1.3 EVELOPMENT OF INTERFACE REQUIREMENTS -- PROJECT OVERVIEW................………………………… 10
2.0 Investigation Of Existing Transit Bus Sub- Systems ................................................... 12
2.1 OBJECTIVES AND PROCEDURE...................................................................................................................... . 12
2.2 POWER STEERING SYSTEM......................................................................................................................... .... 13
2.3 PNEUMATIC BRAKE SYSTEM......................................................................................................................... . 18
2.4 ENGINE/ TRANSMISSION ............................................................................................................................... .. 19
2.5 ELECTRICAL POWER SYSTEM ......................................................................................................................... 21
2.6 CAN IN- VEHICLE NETWORK SYSTEM............................................................................................................. 22
2.7 EXTERIOR GEOMETRY....................................................................................................................... ............. 25
2.8 CONCLUSIONS AND FINDINGS ........................................................................................................................ 29
3.0 Infrastructure/ Vehicle Interface Characteristics.......................................................... 30
3.1 OBJECTIVES AND PROCEDURE...................................................................................................................... . 30
3.2 INFLUENCE ON RUNNING WAY DESIGN......................................................................................................... 30
3.3 INFLUENCE ON STATION DESIGN ................................................................................................................... 31
3.4 VEHICLE EXTERIOR GEOMETRY DESIGN CONSTRAINTS ............................................................................... 32
3.5 INFRASTRUCTURE- BASED LANE TRACKING REFERENCES FOR VAA ........................................................... 33
3.6 CONCLUSIONS AND FINDINGS ........................................................................................................................ 33
4.0 Vehicle Interface Requirements ................................................................................. 35
4.1 VEHICLE SYSTEM ARCHITECTURE ................................................................................................................. 36
4.2 GENERAL DESIGN PHILOSOPHY..................................................................................................................... 40
4.3 OVERVIEW OF IN- VEHICLE DATA COMMUNICATION ................................................................................... 40
4.4 VEHICLE AND LANE POSITION SENSING........................................................................................................ 42
4.5 VEHICLE STATE SENSING........................................................................................................................ ....... 46
4.6 STEERING ACTUATOR ............................................................................................................................... ..... 47
4.7 BRAKE ACTUATOR ............................................................................................................................... .......... 53
4.8 PROPULSION ACTUATOR....................................................................................................................... ......... 55
4.9 CONCLUSIONS AND FINDINGS ........................................................................................................................ 57
5.0 Vehicle Testing And Results...................................................................................... 58
xii
xiii
5.1 OBJECTIVES ............................................................................................................................... ..................... 58
5.2 EXPERIMENTAL PROTOCOLS...................................................................................................................... .... 58
5.3 TEST VEHICLE CONFIGURATION.................................................................................................................. .. 59
5.4 NETWORK TRANSMISSION SPEED AND ITS EFFECTS..................................................................................... 64
5.5 MESSAGE LENGTH......................................................................................................................... ................. 70
5.6 CONCLUSIONS AND FINDINGS ........................................................................................................................ 72
6.0 Conclusions ............................................................................................................... 73
6.1 PROPOSED INFRASTRUCTURE INTERFACE REQUIREMENTS .......................................................................... 73
6.2 PROPOSED VEHICLE INTERFACE REQUIREMENTS.......................................................................................... 74
6.3 NEXT STEPS ............................................................................................................................... ..................... 76
7.0 References ................................................................................................................. 77
8.0 Appendices ................................................................................................................ 78
8.1 APPENDIX A NEW FLYER 40’ CNG J1939 MESSAGE LIST........................................................................... 78
8.2 APPENDIX B NEW FLYER 60’ DIESEL ARTICULATED BUS J1939 MESSAGE LIST ...................................... 80
8.3 APPENDIX C EFFECTS OF TIGHT TURNING RADII ON NEEDED LANE WIDTH ............................................. 82
8.4 APPENDIX D SENSING AND ACTUATING REQUIREMENTS ............................................................................ 86
8.5 APPENDIX E SCHEMATICS OF EXISTING BUS SUB- SYSTEMS........................................................................ 88
8.6 APPENDIX F IMPLEMENTATION OF DATA COMMUNICATION ON TESTING BUS .......................................... 92
xiv
xv
List of Figures
Figure 1.1 Schematic of electric steering actuator hardware………………………………… 4
Figure 1.2 Longitudinal control diagram……………………………………………………….. 5
Figure 1.3 An example of a bus brake actuator………………………………………………… 5
Figure 1.4 Sensing and Actuation for Lateral and Longitudinal Control
using J1939 Serial Network………………………………………………………... 6
Figure 2.1 New Flyer Power Steering Systems……………………………………………… 14
Figure 2.2 New Flyer Steering Column Installation……………………………………………. 14
Figure 2.3 New Flyer steering wheel and column ( left 40’ CNG; right 60’ Diesel)…………... 15
Figure 2.4 Gillig Hydraulic Power Steering System…………………………………………… 15
Figure 2.5 Gillig Steering Wheel and Column………………………………………………….. 16
Figure 2.6 VanHool Steering Wheel and Column ( left 40’, right 60’)………………………… 16
Figure 2.7 Accelerator Pedal Assembly………………………………….……………………. 20
Figure 2.8 J1939 Bus Hardware Interface…………………………………………………….. 24
Figure 2.9 Vehicle Network Configuration for VanHool 40’………………………………… 25
Figure 2.10 Exterior of New Flyer 40’ CNG Bus ( with door extension)……………………... 26
Figure 2.11 New Flyer 40’ CNG Bus Door…………………………………………………… 26
Figure 2.12 Exterior of New Flyer 60’ Diesel………………………………………………… 27
Figure 2.13 Exterior of Gillig 40’ Diesel……………………………………………………… 27
Figure 2.13 Exteriors of VanHool 40’ ( left) and 60’ ( right)…………………………………... 28
Figure 2.14 Opening of VanHool Middle Door............................................................................ 28
Figure 2.15 VAA System Functional Blocks and Information Flow………………………….. 36
Figure 4.1 VAA System: A Modular Distributed System Architecture………………………. 39
Figure 4.2 Schematic for Vehicle Position Sensing…………………………………………… 43
Figure 4.3 Schematic of the Steering Actuator………………………………………………… 48
Figure4.4 Schematic of a General Brake Actuator…………………………………………….. 54
Figure 4.5 Schematic of a General Propulsion Actuator………………………………………. 56
Figure 5.1 Final docking part of PATH Richmond Field Station test track
( dimensions in meters, not shown to scale)………………………………………. 59
Figure 5.2 Test bus hardware configuration…………………………………………………… 60
Figure 5.3 Original PATH VAA System Architecture………………………………………… 61
Figure5.4 Interface Requirement Testing System Architecture………………………………. 62
Figure 5.5 Software Architecture of Sensor/ Actuator Computer……………………………… 63
Figure 5.6 Software Architecture of the Control Computer…………………………………… 64
Figure 5.7 Extended CAN Bus Message Data Frame…………………………………………. 65
Figure 5.8 Network Transmission Speed and Databus Load…………………………………... 66
Figure 5.9 Message Timing for 50 kbit/ sec……………………………………………………. 67
Figure 5.10 Message Timing for 125 kbit/ sec…………………………………………………. 67
Figure 5.11 Message Timing for 250 kbit/ sec…………………………………………………. 68
Figure 5.12 Message Timing for 500 kbit/ sec…………………………………………………. 68
Figure 5.13 Message Timing for 800 kbit/ sec…………………………………………………. 69
Figure 5.14 Docking Performance for Different Network Transmission Speeds…………..…. 70
Figure 5.15 Docking Performance with Different Message Encoding Lengths……………..… 71
Figure 8.1 Additional Lane Width Required vs Turning Radius for a New Flyer 40’ CNG..… 82
Figure 8.2: Additional Lane Width Required vs Turning Radius for a 60’ New Flyer…..….… 82
xvi
xvii
Figure 8.3 Vehicle Motion……………………………………………………………………… 83
Figure 8.4 Turning Radius for Single Unit Bus………………………………………………… 83
Figure 8.5 Offset at Rear Tire for 40’’ Single Unit Bus ( m)……………………………………. 84
Figure 8.6 Turning Radius of Articulated Bus………………………………………………….. 84
Figure 8.7 Offset at Rear Tire for 60’ Articulated Bus ( m)…………………………………….. 85
Figure 8.8 Compressed Air System of New Flyer 40’ CNG Bus……………………………… 88
Figure 8.9 ABS System of New Flyer 40’ CNG bus……………………………………………. 88
Figure 8.10 Compressed Air System of New Flyer 60’ Diesel Bus……………………………. 89
Figure 8.11 Pneumatic Brake System of Gillig 40’ Low Floor………………………………… 90
Figure 8.12 ABS Brake System of Gillig 40’ Low Floor………………………………………. 90
Figure 8.13 Schematic of ABS and ASR System on VanHool Buses………………………….. 91
xviii
xix
List of Tables
Table 2.1 Power Steering System Testing Results……………………………………………… 17
Table 2.2 Summary of existing transit bus sub- systems………………………………………... 29
Table 5.1 Extended CAN message data frame length………………………………………….. 65
Table 8.1 Messages sent from Sensor Computer to Control Computer………………………… 92
Table 8.2 Messages sent from Control Computer to Sensor Computer………………………… 94
xx
xxi
List of Abbreviations
ABS: antilock braking system
ADA: Americans with Disabilities Act
AVCSS: Advanced vehicle control and safety systems
BRT: Bus Rapid Transit
CAN: Controller Area Network
CNG: compressed natural gas
CSMA/ AMP: Carrier Sense Multiple Access / Arbitration by Message Priority
DOP: dilution of precision
EBS: electronic braking system
ECM: electronics control module
ECU: Electronic Controller Unit
FHWA: Federal Highway Administration
FTA: Federal Transit Administration
GCM: governor control module
GPS: Global Positioning System
HMI: Human Machine Interface
INS: Inertial navigation system
JPO: Joint Program Office
OSI: Open System Interconnection
PATH: Partners for Advanced Transit and Highways
PLC: programmable logic controller
ROW: Right of way
SAE: Society of Automotive Engineers
SBAS: Satellite Based Augmentation System
TTCAN: Time- Triggered CAN
TTP: Time- Triggered Protocol
UPS: Uninterrupted power supply
UTC: Coordinated Universal Time
VAA: Vehicle Assist and Automation
VAA- PD: Vehicle Assist and Automation- Precision Docking
VAA- VG: Vehicle Assist and Automation- Vehicle Guidance
VAA- P: Vehicle Assist and Automation- Platooning
VAA- AVO: Vehicle Assist and Automation- Automated Vehicle Operation
VDC: volts of direct current
VRS: Virtual Reference Station ( for differential GPS)
xxii
1
1.0 Introduction
Transit agencies throughout the United States ( U. S.) are facing mounting challenges related to
the provision of high quality and cost effective public transportation solutions for the public.
Transit agencies need to offer convenient and reliable mobility options for customers at a
reasonable cost to the transit agency and locality. Due to the increased cost and constraints on
land use in many metropolitan areas, adding significant lane- miles of roadway is becoming
increasingly difficult. Transportation agencies are investigating means to maximize available
capacity without incurring significant additional costs for new construction. High quality public
transit service should be seen as a viable alternative for regions where congestion is severe and
the potential for significant mode shift could be realized with high quality transit service.
Among the transit options, Bus Rapid Transit ( BRT) is seen as a cost- effective alternative to
more conventional fixed guideway systems that are becoming increasingly expensive to
construct and operate. As current funding ( both federal, state and local) for conventional fixed
guideway transit is becoming more limited, transit agencies have to come up with more cost
effective alternate modes. In the recent development of BRT systems, where new construction
does not take place, new BRT lanes are being carved out within existing ROW constraints. In
2003, Las Vegas re- striped North Las Vegas Boulevard and devoted a lane to transit operations,
while Minneapolis has an ongoing and aggressive program to convert freeway shoulders to
transit- use lanes. Because of the land- use, cost and institutional constraints, BRT- interested
transit agencies have expressed strong desires for technological means that would allow buses to
travel safely on narrow rights of way. The narrow right of way could not only reduce
construction and acquisition costs by as much as 20%, but could also allow for a bike lane or
parking lane on arterial roads. In some cases, a few feet of lane width reduction could affect the
decision whether a dedicated bus lane can be provided.
Lane Assist or Vehicle Assist and Automation ( VAA) systems offer the opportunity of providing
high quality transit service within reduced lane widths. VAA includes four functions that can
transfer portions of the bus driving responsibility from the driver to the VAA system: VAA- PD
provides for precision docking at bus stations, VAA- VG provides for vehicle guidance or
automatic steering on the running way between stations, VAA- P provides for automatic
platooning of buses at close separations and VAA- AVO provides for fully automated vehicle
operations. The VAA- PD function can facilitate passenger boarding and alighting at stations,
while VAA- VG could support reduced lane width, allowing the bus to operate in a designated
lane that is only slightly wider than the bus itself without increasing driver workload. It could be
implemented in 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 initial and primary emphasis in this report is on the VAA- PD and
VAA- VG systems, which are expected to be the first to enter public use. The issues identified
for these systems should in large part be applicable to the more advanced VAA systems as well.
Stakeholders have shown significant interest in VAA. For the transit agency, VAA offers
significant benefits including the delivery of rail- like service, an attractive feature to riders, at a
fraction of rail cost. BRT buses equipped with VAA technologies could provide a similar level
2
of service as conventional fixed guideway systems with the same, if not more, benefits. From
the driver’s perspective, the VAA 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
operation, faster and safer boarding and alighting, better schedule reliability, and increased
mobility for ADA riders.
To address the needs of the transit industry, the U. S. Department of Transportation, through the
Federal Transit Administration ( FTA) and the ITS Joint Program Office ( JPO), have spear-headed
efforts to analyze the impacts that VAA systems would have on bus- based transit systems.
The project, called the VAA Tier II Exploratory project, completed in December 2005, looked at
the potential impacts of VAA technologies on transit operations. The results of this research are
promising, showing that six out of nine typical transit operating scenarios would benefit from
VAA technologies and there is a defined market for VAA technologies. Research and
development on VAA technologies have been conducted for many years. Key VAA technologies
such as lane assist systems have been developed and prototype systems have been developed and
demonstrated. VAA is now being considered as a larger scale demonstration program. Currently,
VAA systems are being marketed towards Bus Rapid Transit ( BRT) systems that are beginning
to operate in the U. S.
One concern that was raised as part of the VAA Tier II Exploratory project was how VAA
technologies could be made commercially available in the United States; current commercially
available VAA technologies are only offered by overseas vendors. Although U. S. research
institutions have developed various VAA technologies and in some cases pioneered the
technology development ( e. g., magnetic guidance and vehicle platooning), none of these
technologies are commercialized yet. There is indeed an urgent need for U. S.- based commercial
VAA systems. Some transit agencies have been looking for foreign VAA products to meet their
immediate needs. However, there are a number of institutional and technological hurdles that
U. S. transit agencies must face in order to deploy imported VAA technologies. The primary
institutional barrier is the Buy America regulations that limit the ability of U. S. transit agencies
using federal funds to acquire VAA technologies sold by non- U. S. companies. The primary
technological barrier is the fact that many of these VAA products are tied to a specific,
specialized and costly vehicle and are difficult to retrofit onto the existing buses produced by
North American bus manufacturers. Whether a U. S.- based or imported VAA system is
considered, the prerequisite is that the VAA system must be able to interface with existing bus
subsystems.
The overall goal of this study is to develop interface requirements allowing VAA systems to be
able to interface with commercially available buses in North America. The project objectives
are:
􀀁 Understand the needs, technical issues and challenges for VAA technologies to interface
with vehicles;
􀀁 Develop interface requirements for both the VAA systems and the vehicles, allowing
maximum compatibility, as well as requirements for the vehicle to roadway infrastructure
interfaces;
3
􀀁 Conduct case studies of the BRT applications for two partner agencies; and
􀀁 Test selected interface requirements using PATH’s test vehicle.
1.1 Needs for VAA Interface Requirements
Transit vehicles in North America are mostly manufactured based on individual transit agencies’
customized requirements. As an industry common practice in the United States, the bus
components such as engine, power steering system and pneumatic brake system are developed
by a variety of suppliers ( e. g. Cummins and Detroit Diesel for engines, Allison for transmissions,
TRW and R. H. Sheppard for power steering system, WABCO and TRW for pneumatic brake
system). The existing bus manufacturing practice is such that different bus manufacturers have
the liberty of using different components provided by different suppliers. Although certain
requirements are established industry- wide, most of the system requirements are driven by
individual designs and component suppliers. On the other hand, several VAA technologies have
been developed by different suppliers ( e. g., guidance systems based on magnetic sensing,
differential global positioning system integrated with inertial navigation systems and video
image processing). Different transit agencies may want to implement one specific VAA
technology or even combine multiple VAA technologies on their selected buses according to
their specific operating conditions and scenarios. The interfaces between VAA components and
the mechanical, electrical and electronic systems on the existing bus, if not defined properly, can
be an impediment to large scale deployment of VAA technologies. Therefore, there is a great
need to understand how VAA systems, based on any practical technology, will interface with
transit vehicles and infrastructure.
In order to facilitate VAA deployment, a standard set of interface requirements will be needed to
allow the suppliers to develop VAA technologies with common interfaces and to allow bus
manufacturers to retrofit the VAA technologies of transit agencies’ choice to different buses
without excessive custom design work or modifications to the existing products. These interface
requirements are very critical to both vehicle manufacturers and suppliers to achieve
compatibility, ease of safety verification/ certification and to lower cost and reduce deployment
time. The standard interfaces are also crucial to the transit operators for maintenance.
Specifically, interface requirements are needed in the following areas:
1.1.1 Electronic Guidance
A VAA- VG or VAA- PD system contains three major components: a set of sensors, actuators
and a processor. Among these components, the steering actuator has the closest interaction with
existing vehicle components.
4
Figure 1.1 Schematic of electric steering actuator hardware
An electric steering actuator design, as shown in Figure 1.1, consists of a steering column, a DC
motor actuating the steering column, an electromagnetic clutch and angle sensors measuring
steering wheel position. Various interfaces exist between the add- on components and the bus
steering column. The DC motor connects to the steering column through a clutch and reduction
gear. An incremental encoder is mounted on the motor shaft to measure the relative position of
the steering wheel. A multi- turn potentiometer is connected with the steering column shaft via
pulley gear and belt to measure the absolute position of the steering wheel. Motor current and
clutch ON/ OFF are controlled by an Electronic Controller Unit ( ECU), which receives a torque
command from an upper level computer and issues corresponding current commands so that the
DC motor will generate the required torque. The clutch can also be controlled by the upper level
computer by issuing a clutch command to the ECU. The ECU has built- in self- diagnostic
features. The health condition of the motor is fed back to the upper level computer through the
motor condition signal. Because these additions can be standardized, interface requirements are
needed to specify the interface between the necessary add- on components and the current
steering mechanism. Additionally, the performance of some of the interface components may
also need to be addressed. For example, some of the existing power assist systems are designed
with excess freeplay, which makes it very difficult to develop a guidance system that will
provide good tracking accuracy. Corresponding to the performance requirements for the
guidance system, there is also a need to define performance requirements to enable the bus
steering mechanism to support the performance requirements of the complete electronic
guidance system.
5
1.1.2 Longitudinal control
Automated longitudinal control, in conjunction with electronic guidance, enables smooth
operation within the BRT lane and high precision stopping at the bus station.
Figure 1.2 Longitudinal control diagram
Figure 1.2 shows a schematic of the longitudinal control data flow. The longitudinal controller
sends a throttle command to the engine and transmission through either the J1939 data bus or
added electronics. Depending on the transmission model, a transmission retarder may or may
not be available for control purposes. Most engine and transmission state information ( e. g.
engine speed, engine torque, torque converter lockup, current gear etc) can be accessed via the J-Bus.
By retrofitting changes to the existing air brake system, the longitudinal controller can send
brake commands to control the air pressure inside the brake chamber. Vehicle states such as
wheel speed and longitudinal acceleration can be available on the J- Bus or from added sensors
( accelerometer).
Figure 1.3 An example of a bus brake actuator
The brake actuation may need to be retrofitted on the existing air brake system. As shown in
Figure 1.3, the control computer sends out brake commands and the proportional valves regulate
the air pressure inside the air brake system according to the received brake command. The most
6
important interface requirements for the brake actuator are how quickly and how accurately the
brake actuator can build up or release the air pressure required by the brake command.
1.1.3 Functions of In- Vehicle Networks
All modern buses use an in- vehicle data network. In buses powered by both Cummins engines
and Detroit Diesel engines, the engine, transmission, and braking systems are all controlled by a
separate Electronic Control Module ( ECM). These ECMs communicate via in- vehicle serial
networks to receive sensing and diagnostic reports and to issue control commands. Most transit
buses use one of three types of in- vehicle networks, namely: SAE J1587, SAE J1922, and SAE
J1939, among which the SAE J1939 network alone can provide the desired data communication
for vehicle control applications. Figure 1.4 shows an example of control functions implemented
using J1939 on New Flyer buses. In order for a VAA system to accomplish assist or automatic
control functions using the existing on- vehicle sensing and actuation functions, communication
through an existing in- vehicle network is essential. However, existing in- vehicle networks
cannot accommodate all VAA communication needs. There is a critical need to develop a
dedicated safety critical in- vehicle network to handle VAA- specific communication needs. This
dedicated VAA in- vehicle network, whether it is implemented using J1939 or other technologies,
must be able to work with the existing in- vehicle network and the VAA subsystems discussed in
this report. Therefore, interface requirements including communication protocols must be
defined.
Figure 1.4 Sensing and Actuation for Lateral and Longitudinal Control using J1939 Serial Network
7
1.1.4 Vehicle - Infrastructure Interface
Certain station/ stop maneuvers, particularly the S- curve docking operation, may not bring the
bus to a stop parallel to the platform. Therefore, the platform may need to take a ‘ non- traditional
design’ in order to accommodate the vehicle trajectory. Also the design of the vehicle may
impact the ability of the vehicle to access the station/ stop, considering features such as the wheel
lugs projecting, the door threshold projection etc. Only when the design of the vehicle and the
configuration of the platform are in harmony, can the buses be precisely docked at the bus stop.
Since station designs will need to be very site specific, a set of design recommendations are
needed.
1.2 Interface Requirements vs. System Requirements and Designs
While the purpose of VAA interface requirements is to define a common set of interfaces
between VAA technologies and the existing and future buses and components, they are not
intended to directly address the system level requirements. However, the VAA interface
requirements need to be consistent and compatible with the potential VAA system designs and to
support VAA system requirements that specify performance, reliability, safety and
maintainability of the system.
Interfaces are closely tied with VAA system designs. There can be various design philosophies
for implementing a VAA system, ranging from a ‘ fully integrated approach’ that requires bus
and VAA components to be designed interactively to achieve maximum integration to the ‘ add-on
approach’ that would design VAA components to fit onto buses from different vendors with
minimum modification of existing bus components. The Phileas bus developed by APTS in the
Netherlands is an example of a fully integrated approach, which involved an integrated design
from the ground up. The automated functions of Phileas were designed in conjunction with the
bus basic driving functions, thereby achieving maximum integration. The integrated approach
enables the physical design and the performance of the basic bus driving functions to better meet
the VAA needs. However, the cost of the integrated approach is extremely high and it is very
difficult to adapt such VAA technologies to existing buses. Additionally, problems can occur if
the VAA functions are too closely coupled with the conventional driving functions. A notable
issue is that failures of the VAA components can affect the basic driving functions. Add- on
designs, though less integrated than the ‘ integrated approach’, support stand alone components to
fit onto existing buses and therefore could likely have wider applications. From the interface
perspective, an integrated VAA system will unlikely require standard interfaces for VAA
components and newly designed buses, whereas it is very important to have standard interfaces
when VAA components and systems are ‘ add- ons’ to existing buses. Therefore, the interface
requirements being studied under this project are specifically for ‘ add- on’ VAA components and
systems. The interfaces would largely rely on existing bus designs and only specify necessary
modifications of the existing systems in order to allow compatibility between the add- on
components and the existing buses and infrastructure.
The VAA system requirements may include system performance specifications and technical
specifications. Collectively, these specifications will define the operation conditions and
environments and will specify the performance, reliability, safety, and maintainability of the
system.
8
The VAA system requirements can impact or be impacted by the VAA interface requirements,
either directly or through system designs. Under the FTA sponsored project ‘ Development of
Needs and Requirements for Transit Lane Assist Systems’, draft performance requirements were
developed. Based on these requirements and prior extensive knowledge of VAA technologies,
this project team established the following considerations for the VAA interface requirements.
1.2.1.1 Interface Requirements vs. Performance Requirements
There are a number of ways that the interface requirements can impact or be impacted by the
overall system performance. For example, a narrower bandwidth in- vehicle network could limit
the update rate of the sensing and control systems, thereby negatively affecting the tracking
accuracy of electronic guidance and longitudinal control systems. The vehicle- infrastructure
interface could also affect how a VAA system performs within the BRT running way and at bus
stations. There is therefore a need to address the interaction between the interface requirements
and performance requirements through analysis or verification tests to validate the impacts of the
performance requirements to ensure that the interface requirements can adequate support a high
performance VAA system.
1.2.1.2 Safety Design Considerations
There is no doubt that all VAA functions are safety- critical. VAA systems may include both
fully automated as well as driver assist functions. In a driver assist system, a driver can become a
portion of the system and could take over the control and be responsible for ultimate safety,
while the fully automated VAA system must be designed to deal with system faults and to
prevent hazardous conditions from occurring. No matter whether VAA involves driver assist or
fully automated operation, it is imperative that the overall system remains fail- safe ( capable of
compensating automatically and safely for a failure) or fail- soft ( capable of operating at a
reduced level of performance and efficiency after the failure of a component or power source) in
the event that a hazardous failure occurs. However, in designing a safety critical system, it is a
common practice that the smallest possible set of safety critical functions are isolated within a
portion of the system in order to reduce the complexity of the overall system design. For a VAA
system, it is imperative that this design philosophy be followed and the safety critical functions
be designed within the VAA components or system, but no fail- safe requirements be imposed on
existing bus components.
Safety requirements are often implemented through redundancy or fail- safe designs. Safety
designs typically involve hardware redundancy and software redundancy. Hardware redundancy
can affect the interface the most if the safety design is propagated to vehicle components. The
assumption is made such that the VAA system would need to work with existing vehicle
components, therefore there is no need for redundant physical interfaces between the add- on
VAA components and the existing vehicle components. On the other hand, the interface
requirements need to support fault diagnosis and software redundancy.
Note that safety design of the VAA requires systematic analyses, which typically would involve
defining the system safety levels, hazard analysis, failure mode effects and criticality analysis,
functional decomposition and identification of safety critical functions. The project team has
9
conducted significant safety analysis of various VAA functions. However, this analysis is not at
the scale that would result in a comprehensive definition of system safety. The project team
recommends a systematic analysis be planned within the upcoming VAA program in order to
address the safety requirements and design issues.
1.2.1.3 Reliability Considerations
Reliability is customarily measured in terms of the mean time between failures ( MTBF) of
infrastructure and onboard systems, subsystems, and components. While it may be technically
possible to build a system that is virtually failure free, after a certain point, the marginal cost for
each additional “ unit” of reliability becomes prohibitive. The reliability of the interface
requirements therefore should be at the same level as the system reliability requirements.
1.2.1.4 Other Design Considerations
VAA systems should be at least as durable as other onboard systems so that the current service
cycle can be maintained ( every 12,000 miles in the case of Lane Transit District, Eugene, OR).
Suppliers of the systems should be required to modularize their system for ease of replacement,
seal them sufficiently to withstand road hazards and bus cleaning, and equip them with a high
level of self diagnostic capabilities. The emphasis should be on a system designed with more
modules rather than fewer. In this way replacement of a module that is beyond repair will be
cheaper, pulling modules and replacing them by the maintenance staff will be easier, and spare
modules will be more like commodity items than specialty items. Interface requirements should
support the modular designs.
At the present time the service life of a bus is approximately 20 years. Given the current pace of
changing technology, the interface requirements may also need to support future upgrades to be
backwards compatible so that the entire system will not have to be replaced.
The requirements and design aspects have been considered in the process of development of the
VAA interface requirements summarized in the final report.
1.2.2 Development of Interface Requirements – Project Overview
In order to begin to address these technological barriers and to facilitate the development and
commercialization of VAA technologies for existing transit vehicles, the FTA and ITS- JPO
funded the research effort to develop interface requirements. A team consisting of Lane Transit
District, AC Transit, California Department of Transportation, and the University of California
PATH Program was selected to develop the interface requirements. Lane Transit District and
AC Transit are members of the BRT Consortium. These agencies have planned dedicated BRT
routes and are convinced that VAA technologies can offer benefits in enhancing the efficiency,
safety and quality of BRT service. Caltrans has been a leading agency supporting development
of advanced technologies for transportation industries, and has devoted significant funding to
sponsor research and development of advanced vehicle control and safety system ( AVCSS)
technologies. Caltrans is interested in supporting the implementation of AVCSS technologies on
transit and other vehicles in order to improve traffic operations and decrease congestion.
10
California PATH, a world- wide leader in the development of advanced vehicle sensing and
control systems, has developed several guidance technologies that have demonstrated superior
performance and practicality for real world deployment. The Gillig, New Flyer and VanHool
bus manufacturers have provided support and information on their bus products. The objectives
were implemented in the project milestones described below.
1.2.3 Assessment of Existing Buses and Supporting Infrastructure Relevant to VAA
As the first step toward the interface requirement definition, a representative selection of existing
transit buses was investigated. Based on extensive knowledge of both VAA systems and the bus
subsystems, the team took a systems approach to investigate bus designs from Gillig, New Flyer,
NABI and VanHool. This investigation focused on the relevant bus subsystems such as steering,
engine/ transmission, brake, in- vehicle network, electrical and exterior geometry, all of which are
directly related to the implementation of a VAA system.
1.2.4 Develop Interface Requirements
In order to clearly identify the system interfaces between VAA subsystems/ elements and bus
subsystems/ elements, a modular system architecture was established. This modular architecture
defines the nature of the interfaces between the VAA system and the bus and bounds the physical
interfaces to a small possible set, therefore is essential for the development of the interface
requirements. Based on the modular VAA architecture, interface requirements were developed.
The interfaces between the VAA system and the transit bus were classified into three major
categories: mechanical, data communication and power supply. Functional units of the VAA
system ( e. g. lane positioning sensing, vehicle state sensing, steering actuator, brake actuator and
propulsion actuator) were analyzed to determine their interface requirements with existing bus
subsystems. Special attention was placed on the data communication, which is the backbone of
the proposed VAA system architecture. The interface requirements between buses equipped
with VAA systems and infrastructure were also studied in terms of running way width, sensed
infrastructure references, boarding platform and vehicle exterior geometry.
1.2.5 Testing of the Requirements
After the interface requirements were developed, a New Flyer bus that was previously
instrumented by PATH with VAA capabilities was modified to emulate the electronic interfaces
between VAA and existing bus controls. The emulated interface was tested on the test track in
order to validate that the proposed interface requirements are technically feasible. A sequence of
verification tests was conducted to quantitatively validate the feasibility of the proposed
requirements.
1.3 Report Organization
This report summarizes the findings on the needs for and the feasibility of interface requirements
based on a thorough study of several transit buses and various VAA technologies. It also defines
the interface requirements for VAA. These requirements are intended as recommendations to
FTA and the transit standard organizations for the development of VAA related interface
11
standards. They can also be used by transit agencies that are considering early deployment of
VAA systems as a basis for developing requirement specifications for the selected VAA systems
and by VAA system suppliers and bus manufacturers as a reference for interface designs in the
absence of official VAA standards. The remaining report is organized as follows:
Section 2.0 reports the results of detailed investigations of buses and component technologies
from several major bus manufacturers and suppliers. Special attention was placed on the
assessment of the suitability of the existing buses as platforms for implementing VAA systems.
Section 3.0 reports the results from an evaluation of bus station configurations, focusing on
designs provided by LTD and AC Transit, and recommended a set of requirements imposed on
the infrastructure for interfaces to a bus equipped with VAA systems.
Section 4.0 reports steps that took place to develop interface requirements for vehicle- borne
system functions. It discusses functional blocks and modular VAA architecture and
recommended VAA interface requirements for in- vehicle VAA systems, including the data
communications between subsystems.
Section 5.0 reports the tests carried out to validate the utility of the interface requirements.
Section 6.0 Conclusions regarding the VAA interface requirements and issues are presented.
12
2.0 Investigation of Existing Transit Bus Sub- Systems
Developing vehicle assist system interface requirements requires extensive knowledge of the
existing transit bus sub- systems that will be interfaced with the new vehicle assist components.
As the first step toward developing the interface requirements, several existing transit buses were
investigated including a New Flyer 40’ CNG, a New Flyer 60’ Diesel, a Gillig 40’ low floor, a
VanHool 40’ and a VanHool 60’. These buses were chosen for two primary reasons. First, they
were readily available to the researchers for detailed study; and second, they have significantly
different characteristics, which together represent a broad range of transit buses ( i. e. single unit
vs. articulated, CNG vs. diesel and domestic manufacturer vs. foreign manufacturer). This
chapter summarizes the current configurations of these buses, focusing on the bus subsystems
that are important for vehicle assist, and describe the tests that were performed on these buses
and the results obtained from those tests. The summary covers the steering system,
engine/ transmission, brake system, in- vehicle network, electrical system and exterior geometry,
all of which are directly related to the implementation of a vehicle assist system. The
information collected here provides an important set of background knowledge, on which the
subsequent requirements definition work can be built.
2.1 Objectives and Procedure
VAA systems are most likely to be designed as add- on or retrofit systems, to be connected to
existing vehicle subsystems. The add- on system design introduces multiple interactions with
existing vehicle subsystems. In order to perform VAA functions such as lane guidance,
precision docking and longitudinal speed control, the VAA system has to access the existing
bus’s subsystems ( e. g. power steering, engine/ transmission and pneumatic brake) to obtain the
desired steering angle, driving force and braking force. The operation of VAA system requires a
variety of real- time information about the operation of the vehicle ( e. g. speed, yaw rate and
steering angle etc). Some of this information is already measured and used by the existing
vehicle subsystems ( e. g. vehicle speed for ABS). Therefore, it is most efficient and cost-effective
to acquire this information from the existing vehicle subsystems without adding new
sensors. As an add- on system, the VAA system also has to draw power ( electrical, hydraulic or
pneumatic) from the existing vehicle power supply and comply with the geometric space
limitations imposed by the existing vehicle design.
Since transit buses are custom- built to meet the requirements of each individual transit agency,
they represent a completely heterogeneous set of characteristics, especially in areas where there
are no existing standards. Due to the diversity of vehicle characteristics and the intense
interactions between VAA system and existing vehicle subsystems, it is essential to gather
information about the key components and sub- systems in current use. The sub- systems and
components that affect vehicle assist functionality include the physical shape and dimensions of
the bus exterior and interior, steering mechanism, engine/ transmission, brake system, data
network and power systems. Field trips were made to a transit agency, a bus manufacturer and
the APTA Expo 2005 to gather information about existing bus subsystems. The effects of the
existing vehicle subsystem designs on the future integration of VAA systems into buses were
assessed based on the information collected from these field trips, the experience with VAA
13
technology implementation in prior PATH experimental projects, and inputs from transit
agencies and bus manufacturers. This information can support the definition of guidelines for
transit agencies and bus manufacturers regarding the design of buses suitable for VAA
retrofitting. At the same time, this information is useful in helping the VAA technology
developers to adapt their technologies to work on the widest possible range of buses.
2.2 Power Steering System
Power steering is a system for reducing the steering effort required of drivers by using an
external power source to assist in turning the wheels. Hydraulic power steering ( HPS) uses
hydraulic pressure supplied by an engine- driven pump, and is popular among heavy duty
vehicles.
The primary focus of the current work is on the vehicle assist functions associated directly with
steering of the bus ( precision docking and vehicle lateral guidance). To keep a bus in a narrow
lane or dock it precisely along a boarding platform, the steering actuator of the VAA system has
to steer the bus’s front wheels to the desired angle using the vehicle’s existing steering system.
Therefore, the characteristics of vehicle’s existing steering system are very important to the
steering actuator design and implementation.
2.2.1 New Flyer 40’ CNG and New Flyer 60’ Diesel
Figure 2.1 shows the power steering system manufactured by R. H. Sheppard for the New Flyer
40’ CNG bus and New Flyer 60’ Diesel articulated bus. It consists of a hydraulic pump, a
reservoir and a power steering gear box. The hydraulic pump supplies pressurized hydraulic
fluid to the hydraulic circuits. When the driver turns the steering wheel, the power steering gear
box will provide enough power assist according to the torque sensed by the torsion bar. Figure
2.2 shows how the steering column assembly is connected to the power steering gear box.
14
Figure 2.1 New Flyer Power Steering Systems
Figure 2.2 New Flyer Steering Column Installation
15
Figure 2.3 shows the actual steering wheel and steering column assembly on the two New Flyer
buses. There is very limited space available for the addition of the steering actuator on the
steering column, as shown in the photos.
Figure 2.3 New Flyer steering wheel and column ( left 40’ CNG; right 60’ Diesel)
2.2.2 Gillig 40’ Low Floor
The steering system of the Gillig 40’ Low Floor bus is manufactured by TRW. Figure 2.4 shows
the hydraulic loop. The steering system consists of the steering wheel, steering column and shaft
assembly, and power steering gear box. The steering column extends through the floor and,
using a universal joint, attaches directly to the input shaft of the power steering box. The power
steering box provides hydraulic power assist when the driver turns the steering wheel. Figure 2.5
shows the actual steering wheel and steering column assembly. There is very limited space for
the addition of the steering actuator as shown in the photos.
Figure 2.4 Gillig Hydraulic Power Steering System
16
Figure 2.5 Gillig Steering Wheel and Column
2.2.3 VanHool 40’ and 60’
Figure 2.6 shows the steering wheel and column installation for both VanHool 40’ and 60’ buses.
There is very limited space for the addition of a steering actuator as shown in the photos. One
particular problem associated with the VanHool 60’ bus is that the trailer wheel will steer with
the front steering wheel when the trailer angle is larger than a certain threshold. This may
become a design constraint for the departure trajectory design of the precision docking maneuver.
Figure 2.6 VanHool Steering Wheel and Column ( left 40’, right 60’)
17
2.2.4 Field Testing of Power Steering System
Since the steering actuator is the most important actuator for the lateral control of VAA functions,
tests were performed to study the static characteristics of the available bus power steering
systems that are relevant to steering actuator design. These systems include a New Flyer 40’
CNG and a New Flyer 60’ diesel bus. Visits took place at the SAMTRANS ( San Mateo County
Transit District) and AC Transit maintenance yards for testing of the Gillig 40’ low floor bus
and VanHool 40’ and 60’ buses.
Tests were performed while the buses are stopped on a paved road with the engine running. A
constant torqueM was applied at the steering wheel to move the bus front wheels at a constant
rotation speed. Tests were also performed to determine the amplitude of the steering mechanism
freeplay. The test results are summarized in Table 2.1.
Table 2.1 Power Steering System Testing Results
New Flyer 40’
CNG Bus
New Flyer 60’
Diesel Bus
VanHool
40’ Bus
VanHool
60’ Bus
Gillig 40’
low floor
Bus
Steering torque
( Nm)
7.9 10.6 4.4 5.2 2.0
Freeplay ( degrees
at steering wheel)
􀀁 25 􀀁 25 􀀁 5 􀀁 5 􀀁 5
Both New Flyer buses are equipped with the same power steering system, manufactured by R. H.
Sheppard, but the New Flyer 60’ Diesel has heavier steering due to its body weight. The existing
power steering gear box has some nonlinear characteristics, which increase the difficulty of
control design for electronic guidance purposes.
􀀁 The power steering gear box has large free play ( about 25 degree steering wheel angle).
􀀁 The power steering gear box cannot provide enough power assist, especially when the
bus speed is low.
􀀁 The power assist curve is generally not symmetric for opposite turning directions ( e. g.
left turn or right turn).
The Gillig 40’ low floor bus’s TRW power steering system is more powerful. It has a smaller
steering wheel ( 0.5 m in diameter compared with 0.56 m for New Flyer buses) and smaller
freeplay in the steering mechanism. The steering torques in Table 2.1 could provide useful
guidance when selecting the components for an add- on steering actuator design ( e. g. the torque
capabilities of the DC motor for the add- on steering actuator design).
The power steering systems on the VanHool 40’ and 60’ buses show similar characteristics to the
Gillig 40’ low floor bus’s TRW power steering system, but with slightly higher steering torque
required.
18
2.2.5 Summary
The design of power steering systems is quite mature, and all the power steering systems
investigated showed similar designs. However, the characteristics of the power steering systems,
especially those related to VAA system functionality differ considerably from system to system.
For example, the power steering systems of both New Flyer buses have significant freeplay,
which makes steering actuator design for a VAA system very difficult. On the contrary, the
power steering systems of both VanHool buses and the Gillig bus have very small freeplay ( less
than 5 degree in steering wheel angle). Secondly, the space available for the installation of a
steering actuator is very limited around the steering columns of all the buses, so a compact
steering actuator design will be necessary.
2.3 Pneumatic Brake System
The pneumatic brake system uses compressed air to deliver the desired braking force at each
wheel’s brakes, and is widely used in heavy duty vehicles. When the driver presses the brake
pedal, the treadle valve is opened and compressed air flows from the air tank to the brake
chambers. The brake chamber is a diaphragm actuator which converts the energy of air pressure
to mechanical force, which is transmitted to the brake pad through the push rod and brake cam.
Brake force is generated by the friction between the brake pad and brake drum. Air is released to
the atmosphere when the driver depresses the brake pedal. The compressor is turned on to
recharge the air tank when the air tank pressure drops below a minimum threshold level.
If longitudinal speed control or longitudinal precision stopping is required for the VAA system,
the brake actuator system has to actuate the bus’s existing pneumatic brake system to deliver the
desired brake force. The addition of a brake actuator to a bus involves a variety of design and
interface considerations that need to be done carefully because any failure in the braking function
would have obviously adverse safety consequences. The brake actuator needs to be able to
engage sufficient braking force to stop the vehicle under the anticipated operating conditions,
while controlling the braking force accurately enough to ensure that the braking action is smooth
enough to promote passenger comfort. This requires thorough knowledge and understanding of
the existing braking systems on the buses that may be equipped for VAA.
2.3.1 New Flyer 40’ CNG
The New Flyer 40’ CNG bus is equipped with an S- cam pneumatic brake system. The on- board
air supply is from an engine- driven air compressor running at engine speed ( Figure 8.8 in
Appendix B). The air system is controlled by a governor. When air pressure in the wet tank is
below 105 psi, a signal is sent to the governor to close the valve in the dryer and start
compressing air into the system. The brakes are applied by depressing the brake foot treadle to
activate the brake foot valve. This causes air to flow from the supply side of the valve to the
delivery side. Once compressed air enters the brake chamber, it drives the S- cam and applies
brake force on the brake drum. Depressing the brake treadle applies a modulated control signal
to the front quick release valve supply port. The quick release valve supplies air through the
ABS modulator (
19
Figure 8.9 in Appendix B) valves which control the supply to the left and right brake chambers
to prevent brake lock- up. Only the rear axle is equipped with an ABS system.
2.3.2 New Flyer 60’ Diesel
The New Flyer 60’ diesel bus is equipped with a similar S- cam pneumatic brake system,
comparable to that on the New Flyer 40’. The on- board air supply is from an engine- driven
Bendix TU- FLO 750 air compressor running at engine speed (
Figure 8.10 in Appendix B). The air system is controlled by a governor. When air pressure in
the wet tank is below 105 psi, a signal is sent to the governor to close the valve in the dryer and
start compressing air into the system. ABS systems are installed on front, center and rear axles.
2.3.3 Gillig 40’ Low Floor
Similar to the New Flyer 40’ CNG bus, the Gillig 40’ low floor bus is equipped with S- cam
pneumatic brake system as shown in ( Figure 8.11 in Appendix B). Both axles are equipped with
an ABS system as shown in ( Figure 8.12)
2.3.4 VanHool 40’
The VanHool 40’ bus is equipped with D- Elsa compressed air disk brakes manufactured by
Lucas. ABS is installed on both front and rear axles. In addition to the ABS system, traction
control such as ASR is also installed. Figure 8.13 shows the location of ABS valves and ASR
valves.
2.3.5 Summary
All the buses investigated are without the centralized electronic control of an electronic brake
system ( EBS). Therefore, actuating the pneumatic brake system through existing in- vehicle
components is not possible. Rather, it is necessary to make more significant modifications to the
existing pneumatic brake system in order to implement a brake actuator.
2.4 Engine/ Transmission
The engine performs the conversion of heat energy generated from fuel burning to mechanical
torque at the drive shaft. Compressed natural gas ( CNG) and diesel fuel are the most popular
fuels used in bus engines, and these engines have significantly different operating characteristics.
To provide a wide range of torque and motion combinations to suit different driving conditions,
the engine torque and rotation are transmitted to the driving wheels of the bus through a variable
speed transmission.
If longitudinal speed control or longitudinal precision stopping is required for the VAA system,
propulsion actuation is needed to actuate the bus’s existing engine/ transmission to deliver the
desired driving force. The characteristics of the engine and transmission must be well
understood before the propulsion controller can be designed, in order to make sure that the speed
and acceleration of the bus can be controlled smoothly and accurately. The smoothness and
20
accuracy are needed for passenger comfort, fuel economy, minimizing emissions and ensuring
safety of the control of the bus motions.
2.4.1 New Flyer 40’ CNG
The New Flyer 40’ CNG bus is equipped with a Cummins C8.3G CNG engine. The Cummins
C8.3G CNG engine is an 8.3 liter, four- stroke, inline, high speed engine. The electronic fueling
control system consists of two separate systems: the electronic control module ( ECM) and
governor control module ( GCM).
A standard electronic accelerator pedal containing a pedal position sensor and idle validation
switch ( as shown in Figure 2.7) is installed to provide inputs to the engine GCM.
The transmission system consists of an Allison B400R transmission, an electronic controller unit
( ECU), the shift selector located on the instrument panel, and a remotely mounted transmission
cooler and accumulator. The ECU receives control inputs from the shift selector and inputs from
the sensors in the transmission control module. It processes this information and sends shift
commands to the control module. The ECU also provides diagnostic information which can be
read as codes through the shift selector or downloaded using a data reader.
Figure 2.7 Accelerator Pedal Assembly
21
2.4.2 New Flyer 60’ Diesel
The New Flyer 60’ diesel bus is equipped with a Series 50 high speed diesel engine
manufactured by Detroit Diesel. The engine- mounted ECM includes control logic to provide
overall engine management. The ECM continuously performs self diagnostic checks and
monitors the other system components. An electronic accelerator pedal similar to that in Figure
2.7 provides the engine ECM with throttle input. The transmission system is an Allison B500R
which is similar to the one used in the 40 foot CNG bus.
2.4.3 Gillig 40’ Low Floor
The Gillig 40’ low floor bus is equipped with a Detroit Diesel Series 50 diesel engine, which is
the same as the New Flyer 60’ diesel. An electronically controlled fueling system is installed.
The electronic fueling system uses an accelerator pedal with an internal potentiometer to regulate
the flow of fuel to the engine. The same transmission ( Allison B400) as that used in the New
Flyer 40’ CNG bus is installed on the Gillig 40’ low floor diesel bus. Details can be referred to
the corresponding section on the New Flyer 40’ CNG bus
2.4.4 VanHool 40’
The VanHool 40’ bus is equipped with a Cummins ISL diesel engine. The accelerator pedal is
connected to the engine fuel injection pump by electric wiring. The electronic controller unit
calculates the desired injected quantity and injection timing for every operating condition. This
results in a reduction of the pollutant emissions and offers an exceptionally high torque at low
engine speeds. The electronic controller unit also monitors the operation of the engine- control
system. When a failure is detected, the electronic controller unit limits the engine speed and
power, and will shut off the engine after approximately 30 seconds. A Voith 864.3 E automatic
transmission with hydraulic retarder is installed.
2.4.5 Summary
Most of today’s heavy duty vehicle engine/ transmissions are equipped with sophisticated
electronic controls. It is very difficult to modify the internal controls of the engine/ transmission
to implement the propulsion actuator design, because these are proprietary to the manufacturers
and are carefully tailored to the specific characteristics of the individual engine/ transmission.
From the propulsion actuator design point of view, it is important to note that most of the
engine/ transmissions are “ throttle- by- wire” systems, using an electronic accelerator pedal. This
makes it easy to provide the propulsion actuation function through an electronic interface with
the VAA control computer, regardless of whether the vehicle is equipped with its own internal
network.
2.5 Electrical Power System
The vehicle electrical power system supplies electrical power to all vehicle subsystems. It
usually includes batteries, which are charged by an alternator driven by the engine. The
22
electronic components of the VAA system need to draw power from vehicle’s existing electrical
power system. In order to minimize power supply complications in implementing the VAA
system, it is desirable to use components that are already compatible with the standard onboard
electrical power characteristics of transit buses.
2.5.1 New Flyer 40’ CNG and 60’ Diesel
The electrical system is a 12/ 24 VDC split system, negatively grounded. All components are
rated at 12 or 24 Volts DC, depending on the system in which they are employed. A PLC
( Programmable Logic Controller) manufactured by Allen- Bradley is used for logical controls.
2.5.2 Gillig 40’ Low Floor
The Gillig 40’ low floor diesel bus has a dual electrical system composed of both 12V and 24V
DC power. The Dinex- MPX multiplex system ( a control system similar to PLC) made by I/ O
Controls Corporation is used to handle logical controls like interlocking of door open/ close, low
air warning, etc.
2.6 CAN In- Vehicle Network System
The Controller Area Network ( CAN) is a serial communication protocol which efficiently
supports distributed real- time control with a very high level of security. CAN was originally
developed by the German company Robert Bosch for use in the car industry to provide a cost-effective
communications bus for in- car electronics and as alternative to expensive and
cumbersome wiring harnesses. The car industry continues to use CAN for an increasing number
of applications, but because of its proven reliability and robustness, CAN is now also being used
in many other industrial control applications. CAN is an international standard and is
documented in ISO 11898 ( for high- speed applications) and ISO 11519 ( for lower- speed
applications)
CAN is a protocol for short messages. Each transmission can carry 0 - 8 bytes of data. This
makes it suitable for transmission of trigger signals and measurement values. It is a
CSMA/ AMP ( Carrier Sense Multiple Access / Arbitration by Message Priority) type of protocol.
Thus the protocol is message oriented and each message has a specific priority according to
which it gains access to the bus in case of simultaneous transmission. An ongoing transmission
is never interrupted. Any node that wants to transmit a message waits until the bus is free and
then starts to send the identifier of its message bit by bit. A zero is dominant over a one and a
node has lost the arbitration when it has written a one but reads a zero on the bus. As soon as a
node has lost the arbitration it stops transmitting but continues reading the bus signals. When the
bus is free again the CAN Controller automatically makes a new attempt to transmit its message.
In the early 90' s, the SAE ( Society of Automotive Engineers) Truck and Bus Control and
Communications Sub- committee started the development of a CAN- based application profile for
in- vehicle communication in heavy duty vehicles. In 1998, the SAE published the J1939 set of
specifications supporting SAE class A, B, and C communication functions. On modern trucks
and buses, the engine, transmission and braking systems are each controlled by separate
Electronic Control Modules ( ECM). These ECMs communicate via in- vehicle serial networks,
23
typically using the SAE J1939 standard. These in- vehicle networks have several important
functions:
􀀁 Broadcast: information about engine speed, wheel speed, current gear and many other
vehicle system states is regularly broadcast by each ECM and may be used by other
ECMs for control or for display of information.
􀀁 Command: the transmission or an anti- locking braking system may command or inhibit
engine speed or torque by sending a message on these networks; advanced cruise control
systems may also use these capabilities. Commands can also be sent to activate
airbrakes, transmission retarders and engine retarders.
􀀁 Fault reporting: special messages report faults. These messages can activate dashboard
" blink code" or error number systems for fault analysis.
􀀁 Off- line diagnostics and information reporting: the in- vehicle networks can be used
for communication with a variety of service tools to report system settings and trip
information, and in some cases can be used to recalibrate the ECM.
The in- vehicle network characteristics are very important to the functionality of VAA systems
for two major reasons. First, the VAA system could tap into the in- vehicle network to acquire
sensor information that is already available on the network. Second, the in- vehicle network
provides a simple channel for the VAA system to actuate the existing vehicle’s
engine/ transmission or brake system if the system configuration allows. Therefore,
understanding the existing in- vehicle networks and integrating the existing in- vehicle networks
into VAA systems could make it possible to simplify VAA system design and save the cost of
additional sensors.
2.6.1 New Flyer 40’ CNG
The New Flyer 40’ CNG bus has a Cummins C8.36+ CM556 electronic control system, which
features both the J1587 ( a slower serial data communications link standard which uses RS- 485
transceivers and receivers) and J1939 serial networks ( Figure 2.8, based on CAN networks). The
transmission, engine and braking system ECMs are all connected together by both the J1939 and
the J1587 serial networks. The J1939 network on this system is configured for a communication
bandwidth of 250Kbps. The engine ECM is not calibrated to respond to the J1939 Torque/ Speed
Control message, and no engine retarder is available. The ABS braking system on the 40 foot
CNG bus is without the centralized electronic control of an electronic brake system ( EBS). Thus
the brake system cannot be controlled via the J1939 network. The detailed J1939 network
messages useful for VAA system can be found in Appendix A.
2.6.2 New Flyer 60’ Diesel
In the New Flyer 60’ diesel bus, transmission, engine and braking systems are all connected by
both J1587 and J1939 networks. The New Flyer 60’ diesel has a Detroit Diesel engine with an
ECM that broadcasts on both J1587 and J1939 networks, and also responds to J1939 Torque/
Speed Control command requests for engine torque and engine speed. No engine retarder is
configured, and engine retarder messages sent to the engine ECM are ignored. The ABS braking
24
systems on the 60 foot articulated bus are without the centralized electronic control of an EBS
system. Thus the brake system cannot be controlled via the J1939 network. The detailed J1939
network messages useful for VAA system design can be found in Appendix B.
Figure 2.8 J1939 Bus Hardware Interface
2.6.3 VanHool 40’
The VanHool bus uses a KIBES multiplex system. The vehicle network is composed of 5 CAN
buses which connect dashboard, transmission ECU, door controls, ABS/ ASR and engine ECU
( Figure 2.9). A central computer node serves as both bus gateway between different CAN buses
and logic control unit. The messages are compatible with the J1939 standard.
2.6.4 Summary
With the increasing complexity of onboard electronics, the in- vehicle communication network
becomes more important. Although the SAE J1939 standard was developed to serve as the in-vehicle
communication network for heavy- duty vehicles, the exact information available from
the in- vehicle communication network can vary significantly for different vehicles. Furthermore,
the configuration of different transit buses’ subsystems is also different with respect to the in-vehicle
communication network. For example, the engine of the New Flyer 60’ Diesel bus is
configured to react to commands from the in- vehicle communication network, while the engine
25
of the New Flyer 40’ CNG bus does not respond to commands from the in- vehicle
communication network. This diversity of characteristics will require diversity in the designs of
the VAA systems for different buses.
Figure 2.9 Vehicle Network Configuration for VanHool 40’
2.7 Exterior Geometry
To take advantage of the precision docking function of VAA, the bus exterior geometry design
has to be subjected to certain design constraints. These are important in order to ensure that the
bus can approach close enough to the loading platforms at the bus stations without encountering
mechanical interference between any parts of the bus and the platforms or curbs. An
investigation of current bus exterior geometry designs and their implications for close approach
to loading platforms is an important first step toward the development of requirement for bus
exterior geometry.
2.7.1 New Flyer 40’ CNG
As shown in
Figure 2.10, the New Flyer 40’ CNG bus body includes a raised rubber wheel fender around each
wheel. Since these wheel fenders define the outside perimeter of the bus and the actual door step
26
of the existing design are within the vertical plane of the bus outside perimeter, a large gap
between the door step and the curb will be created when the bus docks at a station. It is
recommended that modification be made to extend the door step so that the edges of the door
extension ( Figure 2.11) can be aligned with the outermost perimeter of the bus body, to the
extent this can be done without violating legal limitations on the bus width.
Figure 2.10 Exterior of New Flyer 40’ CNG Bus ( with door extension)
Figure 2.11 New Flyer 40’ CNG Bus Door
2.7.2 New Flyer 60’ Diesel
Figure 2.12 shows the exterior of the New Flyer 60’ diesel articulated bus. Similar to the New
Flyer 40’ CNG bus, the raised rubber wheel fender ( as in
27
Figure 2.12) and inward door step will create a large gap when the bus docks at a station.
Figure 2.12 Exterior of New Flyer 60’ Diesel
2.7.3 Gillig 40’ Low Floor
As shown in
Figure 2.13, a similar raised rubber wheel fender appears on the Gillig bus, which is likely to
create a large gap when the bus docks at the station.
Figure 2.13 Exterior of Gillig 40’ Diesel
2.7.4 VanHool 40’ and 60’
28
The exteriors of the VanHool buses are very clean designed and well suited for the precision
docking function ( Figure 2.14). One of minor problems is the opening of the bus’s middle door.
When it is opening, the door moves toward the outside first and then slides aside as shown in
Figure 2.15. If docking precisely at the platform, the rubber seal at the bottom of the door will
be stuck between the boarding platform and vehicle body, thereby stopping the opening of the
middle door.
Figure 2.14 Exteriors of VanHool 40’ ( left) and 60’ ( right)
29
Figure 2.15 Opening of VanHool Middle Door
2.7.5 Summary
Most of today’s buses’ exterior designs are not ready for precision docking. New designs or at
least some cosmetic changes are needed in order to ensure that the buses are able to approach the
docking platforms and open and close their doors without interfering with the platform.
2.8 Conclusions and Findings
Table 2.2 presents a summary of the key findings from the review of the VAA- relevant
characteristics of existing transit buses. The fundamental conclusions from this review are:
􀀁 The designs of basic bus subsystems such as power steering, engine/ transmission,
pneumatic brake system and in- vehicle network are quite mature and similar for the buses
from the different bus manufacturers.
􀀁 Although the designs of subsystems are similar, the characteristics of each specific
subsystem are quite different ( e. g. the characteristics of power steering system, the
available information from in- vehicle communication network and the capability of
actuating existing vehicle subsystems through the in- vehicle communication network).
To deal with such broad diversity, VAA system designs have to be flexible.
􀀁 None of the buses investigated is completely suitable and ready for implementation of the
VAA system application. Some of the existing subsystems ( e. g. brake system and
exterior geometry) will need modifications to adapt to the VAA system application needs.
Table 2.2 Summary of existing transit bus sub- systems
30
3.0 Infrastructure/ Vehicle Interface Characteristics
VAA systems involve close interaction between vehicles and the infrastructure, so the interfaces
between the vehicles and infrastructure are critically important to the successful design and
implementation of VAA systems. The scope of this project is focused on definition of interfaces,
so when infrastructure issues are discussed, specific attention is given to addressing the
interfaces with the vehicles providing VAA services, rather than the much broader issues of
infrastructure design. There are two primary aspects of these interfaces to consider. First,
modifications of existing infrastructure designs such as running way and station platforms are
required to take full advantage of VAA functionalities. Second, accurate and robust
determination of vehicle position with respect to lane center is critical to VAA system
performance, and must depend on one or another form of infrastructure- based reference support.
3.1 Objectives and Procedure
The roadway and bus stops may need to be constructed or modified based on VAA requirements,
the chosen technology, and selected services. These construction requirements may include the
strength distribution of the pavement, the smoothness of the curvature and its transition
characteristics, the accuracy of the width of the roadway, as well as the smoothness and the
precision of the bus stop locations. The requirements may be site dependent and need to be
determined in the deployment phase. These are generally not mandatory prerequisites for the use
of VAA systems, but they can significantly affect the effectiveness of the systems in practice.
Depending on the technology choice for the specific VAA system, certain infrastructure
references ( specific lane marking or striping, magnetic markers, wires, mechanical guide,
electronic map, or differential GPS signals) are needed to support the specific sensing system.
The sensor and the installation of the reference determine the accuracy of the lateral
measurements. The “ smoothness” of the road reference defined by such infrastructure
significantly influences the ride quality when high tracking accuracy is required. Installation
requirements such as accuracy and smoothness will need to be determined to ensure the
performance of the guidance system as a whole.
In this chapter, vehicle- infrastructure interface requirements will be covered in areas such as
running way design, station design, infrastructure- based reference and vehicle exterior geometry.
The analyses are based on vehicle kinematics and PATH’s prior experience in the development
and implementation of experimental VAA systems.
3.2 Influence on Running Way Design
Running way design has a direct influence on the ability of VAA systems to provide benefits
such as promoting a rail- like image or allowing vehicles to operate in narrower lanes. While a
rail- like image could be achieved through other design elements, enabling buses to operate in
narrow lanes requires the use of vehicle assist technologies. Buses equipped with VAA systems
could operate on narrower lanes than a normal bus can tolerate. The minimum running way
width is determined by the following factors:
31
􀀁 Bus width: Obviously, the width of running way has to be larger than the bus width.
􀀁 Tracking accuracy: The lateral position tracking accuracy of the VAA system determines
how accurately the bus could operate within the designated lane boundary.
􀀁 Design factor: The design factor has to be determined by the system designer based on
the system reliability, operational speeds, the effectiveness of the fault management
system and environmental conditions such as weather.
􀀁 Minimum clearance: Minimum clearance defines the required lateral clearance between
the side of the bus and the lane boundary. This may need to include allowance for a
“ shying distance” to accommodate the responses of drivers of vehicles traveling in the
opposite direction in adjacent lanes
􀀁 Curvature offset: As explained in Section 8.3 “ Effects of Tight Turning Radii on Needed
Lane Width,” the sharper the curve and the longer the vehicle wheelbase, the wider the
lane needs to be in curves. The curvature offset can be determined using the same
equation as that for the offset tracking in Section 8.3 if the bus does not have rear- wheel
steering.
The minimum lane width needed on relatively straight sections of running way is the sum of the
following factors:
Bus width +
Design factor * ( tracking accuracy + minimum clearance) * 2
The following example illustrates the calculation of minimum lane width. For a 7.62 cm ( 0.25
ft) VAA system tracking accuracy, 2.59 m ( 8.5 ft) bus width, 7.62 cm ( 0.25 ft) minimum
clearance, straight road, and design factor of 2, the resultant lane width would be 3.2 m ( 10.5 ft).
Other issues involving running way design include:
􀀁 Strength distribution of pavement: Buses equipped with VAA systems follow their
trajectory with high accuracy and repeatability. A rail- like pavement ( i. e., two narrow
pavements located under the tires of the bus) may be preferred for the pavement design,
but this pavement will have to be designed for more concentrated loading than
conventional pavements.
􀀁 Smoothness of the curvature and its transition characteristics: Since the VAA bus
follows the predetermined trajectory closely; the smoothness of the curvature and its
transition characteristics will contribute significantly to the smoothness of bus motion
and passenger comfort. The curves and their transition spirals therefore need to be
defined precisely based on the levels of lateral acceleration and jerk that correspond to
the desired passenger ride quality.
3.3 Influence on Station Design
Traditional station designs need to be modified to accommodate the requirements of precision
docking. The influence of precision docking on the station design can be summarized as
follows:
32
􀀁 Boarding platform geometric constraint:
o Boarding platform height: According to the Americans with Disability Act ( ADA)
requirements, the vertical gap between vehicle floor and station floor shall be within
plus and minus 1.58 cm ( 5/ 8 in). Most modern buses are equipped with air
suspensions, so the bus height may change due to load variations. Therefore, such a
stringent requirement may not be realistic for the bus docking design, especially
when combined with the effects of a crowned road surface.
o Boarding platform orientation: Another consideration for the boarding platform is the
bus alignment when stopped. Due to the constraints of bus kinematics, it is very
difficult to align both the front bus door and rear bus door to a straight boarding
platform simultaneously, with the same lateral gap for S- curve docking ( docking
involving a lane change), especially for buses with longer wheelbases, such as
articulated buses. This inevitably creates a larger lateral gap at the rear door when S-curve
docking is performed. To solve this problem, the boarding platform should be
designed to tilt toward the bus rear door.
􀀁 Entrance/ exit running way design:
o Entrance running way design: To align both front bus door and rear bus door
simultaneously to the boarding platform with the same lateral gap, a straight
entrance running design is required for the in- line docking scenario. The length of
straight entrance running way is determined by the wheel base of the bus and the
performance of the VAA system. For the S- curve docking scenario, a “ swing- in”
curved road can be designed before the bus docks to the platform.
o Exit running way design: Attention should also be paid to the design of the
departure curve for articulated buses with trailer wheel steering. The initial
departure angle should not be too large so that it does not trigger trailer wheel
steering and rear wheel will not touch curb. The exit path also needs to account for
the overhang of the rear of the bus body behind the rear axle, and its potential to
swing wider than the rear wheel trajectory.
In summary, the two most important design elements for stations using precision docking are:
1) the vehicle floor height and boarding platform floor height need to be equal
2) the entrance/ exit running way needs to be as straight as possible
3.4 Vehicle Exterior Geometry Design Constraints
As shown in Section 2.7, traditional bus body design usually includes raised rubber wheel
fenders around each wheel. Since these raised rubber fenders define the outside perimeter of the
bus and the actual door step may be inboard of the vertical plane of the bus’ outside perimeter, a
gap between the door step and the curb may be created when the bus docks at a station. It is
recommended that in this case modifications be made to extend the doorstep so that the edges of
the door extension can be aligned with the outermost perimeter of the bus body, subject to legal
constraints on the total vehicle width. If wheel fenders extend beyond this width, they should be
truncated at the level of the loading platform in order to avoid interference. Bus doors should be
33
designed that they can still open without interference when docking at a platform that is at the
same height as the bus floor.
3.5 Infrastructure- Based Lane Tracking References for VAA
Determining the vehicle’s lateral deviation relative to the lane center with high accuracy, high
bandwidth and robustness is very important to the successful implementation of electronic
guidance/ assist systems. All lateral guidance technologies require infrastructure- based reference
support of one type or another. The specific requirements for the infrastructure- based references
are determined by the selected sensing technology.
􀀁 Magnet reference system: A magnet sensing system uses magnetic material ( e. g.,
magnetic tape or discrete markers) located on, or embedded in the lane center. If
discrete magnetic markers are used, the distance between magnetic markers cannot be
too long; one meter apart is a good start for precision docking applications. Magnetic
markers, if not installed properly ( e. g., buried too deep below the road surface, or not
perpendicular to the road surface), may increase the noise effect on lateral position
estimation.
􀀁 GPS reference system: To meet the vehicle position sensing accuracy requirement of
the VAA system, the differential GPS technique is usually employed. The differential
correction signal can be made available through several different avenues. Base stations
can be established in the interested area and the differential correction signal is then
broadcast through a radio link with added infrastructure cost. The location of the base
station should be optimized for the signal availability throughout the bus route. The
differential correction signal can also be available through the Satellite Based
Augmentation System ( SBAS) ( e. g. paid services such as StarFire and OmniStar) and
web- based Virtual Reference Station ( VRS). Digital maps are also part of the sensing
infrastructure for the GPS sensing system. The digital map should be detailed enough to
provide the required accuracy and must allow access and map calculations to meet the
real- time requirement.
􀀁 Vision reference system: The markings painted on the road surface for the vision
sensing system must be visible with sufficient contrast under all intended operating
conditions and able to last under the ambient environmental conditions with a
reasonable investment in maintenance.
3.6 Conclusions and Findings
This chapter has addressed the interactions between the vehicle and infrastructure designs, and in
particular the ways in which infrastructure designs need to be adapted to support VAA
implementation. The primary issues for the infrastructure are:
􀀁 Running way: The main influence on running way design is focused on the running way
width, which can potentially be reduced significantly below standard lane width ( 12 feet
for most cases). Other design factors for the running way such as pavement design and
curve design are also discussed.
34
􀀁 Stations: Boarding platforms and entrance/ exit profiles may have to be modified to
accommodate the requirements of precision docking.
􀀁 Vehicle exterior geometry: Precision docking imposes design constraints on the vehicle
exterior geometry compared with traditional bus body design, in order to enable the bus
to approach the boarding platform very closely.
􀀁 Infrastructure- based lane tracking references: Infrastructure- based references are integral
parts of the most important VAA subsystem: vehicle and lane position sensing. All
lateral guidance technologies require the support of infrastructure- based reference
information in one form or another. The differing requirements on the infrastructure-based
references are discussed for the primary existing lateral guidance technologies.
35
4.0 Vehicle Interface Requirements
VAA systems are composed of add- on functional units interacting with the basic bus subsystems
normally controlled by the bus driver. Specifically,
􀀁 The VAA lateral control system will interact with the steering system;
􀀁 The VAA longitudinal control will interact with the engine/ transmission system and the
pneumatic braking system;
􀀁 The VAA system may need to take data from and send data to the existing data bus;
􀀁 Electronics for the VAA system will need to be powered by the bus electrical system.
The objective is to ensure that VAA systems designed with different technologies can interface
seamlessly with buses manufactured by different bus manufacturers. Through the review of
vehicle and infrastructure elements, it is evident that buses in North America use components
from different suppliers, and therefore have discrepancies in respective designs. However, it was
found that common vehicle interfaces can be defined for VAA subsystems to interact with the
existing bus subsystems. This important finding establishes the foundation for a standard set of
interface requirements that can be adopted by all manufacturers.
In order to define the VAA interface requirements, it is necessary to establish a VAA
architecture through which the interface between VAA subsystems and other bus subsystems can
be clearly identified. This VAA architecture needs to be modular such that interactions between
VAA subsystems and other bus subsystems are streamlined. These interfaces can be defined to
support all VAA performance requirements, without becoming unnecessarily complicated or
burdensome. Based on the VAA architecture, interactions between the VAA subsystems and
existing bus subsystems are analyzed. Following this design philosophy, the following design
steps are taken in order to develop the VAA interface requirements:
i. A modular VAA system architecture is defined. The interfaces between VAA
subsystems and existing vehicle subsystems are identified.
ii. The interfaces are classified into three categories -- mechanical interface, power supply,
and data communication. The general design philosophy is then introduced.
iii. Data communication is more challenging than the other two interface categories.
Therefore, an introduction to data communication is presented and the shared in- vehicle
network is introduced in detail as the backbone of the modular system architecture.
iv. Because of the complexity of the VAA system, a “ divide and conquer” design method is
employed ( i. e., the design is carried out for each VAA system functional block in each
category). The emphasis is placed on important functional blocks such as vehicle and
lane position sensing and steering actuation.
36
4.1 Vehicle System Architecture
4.1.1 Functional Blocks and Information Flows
As the first step of system design, the VAA system is partitioned into several functional blocks.
Detailed analysis will be presented to show how the functional blocks interact with each other
and how they interact with the existing bus subsystems. Figure 4.1 shows the schematic of a
VAA system, organized by functional blocks, with information flows shown between functional
blocks and its interactions with the driver, the existing bus subsystems and the infrastructure.
The VAA system is composed of the following functional blocks:
􀀁 Sensing/ Communication: Sensing directly interacts with existing bus components and
with external infrastructure support to provide information on vehicle states and position.
Information can also be exchanged between the vehicle and roadside and among different
vehicles through wireless communication.
o Vehicle state sensing: The components in this category potentially consist of
existing or additional vehicle sensors. The vehicle state information includes
vehicle speed, engine speed, gear position and door opening, etc. It provides
necessary information for controller and fault detection/ management.
o Vehicle position sensing: Through the interaction with sensor reference
infrastructure, vehicle position sensing detects the vehicle position with respect to
the lane boundary. It is the key sensor in the VAA- PD and VAA- VG systems.
Vehicle position sensing systems using computer vision, magnetic sensing,
mechanical contact and Global Positioning System ( GPS) sensing all require
infrastructure support of some sort.
o Communication: Communication includes vehicle- to- vehicle communication
( e. g. to support VAA functions such as platooning) and roadside- to- vehicle
communication ( e. g. acquisition of differential signals for Differential GPS).
Figure 4.1 VAA System Functional Blocks and Information Flows
37
􀀁 Actuating: Actuating directly interfaces with the bus's power steering system, pneumatic
brake system and engine/ transmission to provide steering, driving and braking force for
the desired vehicle maneuvers.
o Steering actuator: The steering actuator receives control commands from an
upper level controller and actuates the existing steering system to the desired
steering angle. This is the key actuator in a VAA- PD or VAA- VG system. It can
also be used as a haptic device, providing torque feedback to alert the driver.
o Brake actuator: The brake actuator receives control commands from an upper
level controller and actuates the existing bus’s pneumatic brake system to provide
desired braking force.
o Propulsion actuator: The propulsion actuator receives control commands from
an upper level controller and actuates the existing engine/ transmission to provide
the desired driving force.
􀀁 Controller: The controller is the brain of the VAA system. It receives commands from
the driver through the Human Machine Interface ( HMI) and relevant sensing information
from the sensing systems. Appropriate commands are then calculated and sent to the
actuators to achieve the desired maneuvers.
o Lateral controller: The lateral controller calculates the steering command that is
sent to the steering actuator according to the received sensor information so that
the bus stays within the lane boundary or close to the docking platform.
o Longitudinal controller: The longitudinal controller calculates the braking and
propulsion commands sent to the brake and propulsion actuators so that the bus
maintains the desired speed or stops at the exact location for precision docking.
o Coordination controller: The coordination controller issues commands to both
lateral controller and longitudinal controller to achieve the desired bus maneuvers
( e. g. lane keeping or precision docking)
􀀁 Human machine interface ( HMI): The HMI is the bridge or communication channel
between the driver and the VAA system. It can serve multiple functions, including
providing diagnostics, warnings, driver assistance, system activation or deactivation via
multiple modalities ( audible, visual, or haptic feedback to driver).
􀀁 Fault detection and management: Fault detection and management form a necessary
functional block for the VAA system because it is a safety critical system. Alerts will be
issued to the driver when failures and inconsistencies are detected in sensor, actuator or
controller functioning. The VAA system will then operate in a failure mode with
degraded performance but guaranteed safety.
􀀁 Infrastructure: A VAA system includes the special characteristics of the lanes
themselves, which may include dedicated lanes and docking platforms as well as visual
or magnetic lane markings for sensing. Installing a VAA system may include lane
construction, sensing infrastructure installation, platform construction and roadside
communication link.
38
Generally, the VAA system operates as follows:
􀀁 The sensing/ communication block obtains information such as vehicle lane position and
related vehicle states ( e. g. vehicle speed, brake pressure etc) from its interactions with
sensing infrastructure, data communication with existing bus subsystems, other VAA
subsystems and wireless communication with other buses and the roadside .
􀀁 The acquired sensing information is made available to the controller, HMI and fault
management subsystems through data communication.
􀀁 Once the controller receives such information, control commands ( such as steering,
propulsion and braking) will be calculated and sent to the corresponding actuators.
􀀁 The actuators actuate existing bus subsystems such as power steering system,
engine/ transmission and pneumatic brake system according to the received commands so
that the desired vehicle maneuver ( e. g. lane keeping with certain cruising speed and
precision docking) is achieved.
􀀁 The bus driver monitors and controls the VAA system activation through the HMI.
In Figure 4.1, the green blocks indicate VAA subsystems onboard the bus. These blocks
exchange information through data communication. The light blue block at the right of Figure
4.1 represents the interface to the infrastructure, which is discussed in Section 3.0.
4.1.2 System Architecture Design
The goal of system architecture design is to provide an architecture that can incorporate different
VAA technologies and interface with different existing bus subsystems without fundamental
changes. In this section, the advantages of a modular system architecture connected by a shared
in- vehicle network are discussed first, followed by introduction of the recommended VAA
system architecture.
Compared with centralized system architecture, a modular system architecture with a shared
network has important advantages. Today's vehicles contain hundreds of circuits and sensors,
and many other electrical components. Communication is needed among the many circuits and
functions of the vehicle. In early vehicle systems this type of communication was handled via a
dedicated wire through point- to- point connections. If all possible combinations of switches,
sensors, motors, and other electrical devices are accumulated, the resulting number of
connections and amount of dedicated wiring would be enormous. A modular system architecture
with a shared communication network provides a cheaper, safer, more reliable and efficient
solution for today's complex vehicle systems.
In- vehicle networking, also known as multiplexing, is a method for transferring data among
distributed electronic modules via a shared data bus. Without shared in- vehicle networking,
inter- module communication would require dedicated, point- to- point wiring, resulting in bulky,
expensive, complex, and difficult to install wiring harnesses. Applying a shared data bus reduces
the number of wires by combining the signals on a shared data bus through time division
multiplexing. Information is sent to individual control modules that control each function, such
as anti- lock braking, turn signals, and dashboard displays.
39
Figure 4.2 VAA System: A Modular Distributed System Architecture
As shown at the left side of Figure 4.2, a modular system architecture with shared in- vehicle
network has already become common practice in industry. For heavy duty vehicles such as
trucks and buses, the SAE J1939 standard has been developed as the standard of communication
protocols among electronic controller units. Standardized protocols allow modules from many
suppliers to easily link together, forming a type of “ open architecture”. Such a flexible “ open
architecture” allows easy expansion of system function with additional new modules. Another
trend in the vehicle system architecture is that multiple data buses are implemented in vehicles.
Usually, a low speed data bus is used for locks, windows and other devices. A high speed one is
used to connect devices such as engine/ transmission, ABS and traction control, which are related
to real time control ( e. g. VanHool’s in- vehicle network architecture as shown in Figure 2.9).
To comply with the common industry practice, especially for heavy- duty vehicles, the VAA
system architecture is designed as a modular system connected by a dedicated VAA data bus as
shown in Figure 4.2. The dedicated data bus for the VAA system is connected to the existing
vehicle data bus through a bus gateway. Functional blocks such as sensing, actuators and
controllers are designed as smart modules that are connected by the dedicated data bus. In order
to allow modules ( e. g. steering actuator and vehicle position sensing) from different suppliers
with different technologies to be combined in a VAA system, interface requirements between the
VAA sub- systems and existing bus sub- systems should be established. The interfaces between
VAA system modules ( i. e., the communication protocol for the dedicated data bus) should also
be specified.
From the existing transit bus system point of view, engine/ transmission, brake, power steering
and existing CAN bus are the sub- systems that directly interface with the VAA system. They
were already discussed in Section 2.0. From the VAA system point of view, the functional
40
blocks that directly interface to the existing transit bus subsystems are sensing and actuating.
These will be discussed in detail in the rest of Section 4, along with data communication between
different functional blocks, which is an integral part of the VAA system. From the infrastructure
point of view, lane construction, boarding platform and sensing infrastructure directly interface
with the VAA system. They were already discussed in Section 3.0.
Based on the knowledge of the existing vehicle subsystems and infrastructure reviewed in
Sections 2.0 and 3.0, the interface requirements are described here for VAA system functional
blocks and existing vehicle subsystems. The objective is to provide a unified set of interface
requirements to accommodate a full range of VAA system technologies and existing vehicle
subsystems.
4.2 General Design Philosophy
The interfaces between the VAA system and an existing transit bus can be classified into three
major categories: mechanical ( including mechanical installation, hydraulic or pneumatic
connections), data communication ( including dedicated signal connections and in- vehicle data
networking) and power supply ( including electrical, hydraulic and pneumatic power). Although
the exact interface requirements are subject to the specific system design, general guidelines for
VAA system interface requirements are:
􀀁 The design and implementation of the VAA system shall not affect normal manual
driving operations.
􀀁 The design and implementation of the VAA system shall not interfere with existing
vehicle components mechanically, electronically or electro- magnetically so that it will
not imperil or degrade performance of existing vehicle components and systems. For
example, the electric power consumed by the VAA system shall be calculated carefully.
If the consumed power is too large, a larger alternator may be needed to ensure smooth
operation of the existing bus systems.
􀀁 The design and implementation of the VAA system shall tolerate normal wear and tear of
any related or connected bus components.
􀀁 The implementation and application of the VAA system shall not jeopardize existing and
new safety- critical operations.
4.3 Overview of In- vehicle Data Communication
Data communication can be implemented as point- to- point signal connections, a shared data
network or various combinations of both types of communication. To ensure a simple, modular,
expandable, upgradeable, reliable and redundant design for safety concerns, a shared data
network approach is preferred as shown in Figure 4.2. In such a configuration, individual
functional blocks such as sensors, actuators, HMI and controller have their own local processors.
These “ smart” functional blocks communicate via a common data bus to form a distributed real-time
control system. The data communication network subsystem functions as the backbone for
the distributed system and becomes a critical component. From the multi- layered network Open
41
System Interconnection ( OSI) model point of view, the data communication network subsystem
can be segmented into several different layers. The focus of this report is on the application
layer. What are the necessary messages exchanged among the different functional blocks of the
VAA system? How often will these messages be exchanged? What is the priority of each
message? These are the questions to be addressed here. The answers support the definition of
the message framework as well as information interface requirements.
How to implement the data communication network subsystem in the lower level of the network
OSI model is not the focus of this report, since this is subject to change with advances in
technology. Different communication protocols have been proposed for distributed real- time
control systems, especially for the safety- critical automotive applications such as X- by- wire ( X
means steering, braking or throttle). The Controller Area Network ( CAN) is a serial
communications protocol that supports distributed real- time control applications with
dependability requirements. CAN networks have the characteristic that the highest priority
message active on a CAN network is always delivered, regardless of conflicting messages. CAN
is popular in automotive electronics such as engine control modules, transmission control
modules, and Anti- lock Brake Systems ( ABS) with bit rates up to 1Mbits/ sec. The SAE J1939
protocol is a vehicle application layer built on top of the CAN protocol and is currently a widely
implemented standard for heavy- duty vehicles including transit buses. J1939 has already defined
messages at the vehicle application layer level for common powertrain ( engine, transmission and
braking) applications. The J1939 Torque/ Speed Control message already provides much of the
information required by the longitudinal controller. There is still a great deal of undefined
message space in the J1939 standard available for use by future applications areas, one of which
could be VAA. In the short term, the proprietary message space can be used to implement
messages supporting VAA functionality.
A major drawback for CAN protocol implementation of distributed real- time systems is that
CAN is an event- triggered communication protocol and requires careful analysis of the relative
priorities and frequencies of all messages on the network in order to guarantee the timely
delivery of messages required by real- time control systems. Several different protocols ( e. g.
FlexRay, SAFEbus, Time Triggered CAN ( TTCAN), and Time- Triggered Protocol ( TTP)) have
been proposed to add the time triggered communication and other functions suited for real- time
control systems. However, these proposed communication protocols are not yet widely
implemented in the heavy vehicle market.
4.3.1 Message Types
In general, messages exchanged between different functional blocks can be classified into the
following categories:
􀀁 Identification: Identification or source address is the unique signature for each electronic
controller unit that sends the message. It could include component ID not only for the
components of different functional blocks but also for the components of the same type
of functional blocks when redundancy is used to address reliability.
􀀁 Status: When a distributed real- time system configuration is utilized for safety- critical
control functions, it is important that all the functional blocks connected together share a
42
common view of the system state and use the same system state to compute outputs. To
achieve synchronization among functional blocks, periodic message passing system and
component status can be introduced. This status includes component status ( e. g.
ready/ not ready and normal/ fault) and operation status ( e. g. acknowledgement of message
receipt and the resulting status for certain operation such as calibration, control and
manual/ automatic transition).
􀀁 Command: Commands can be issued by certain functional blocks to other functional
blocks such that certain operations will be performed or certain information will be
provided.
􀀁 Health signal: A health signal is a specialized status message. It does not provide the
sender’s status directly. With such a signal, other functional blocks could diagnose the
sender’s status. It could be a heart beat signal or a continuous counter embedded in a
message.
􀀁 Data: Most of the traffic on the data communication network is data exchanged between
functional blocks. It could be the sensor measuring results, parameters for certain
functional blocks’ operations, and commands.
􀀁 Redundant Message: One way to improve system reliability of the data communication
network is redundant message passing. The redundant message could be a simple
replica of the original message or the original message with different encoding.
4.3.2 Message Properties
􀀁 Update method: Updates for sensor or status parameters can be broadcast on the
network periodically or supplied only in response to queries from other functional blocks.
􀀁 Update frequency: The update frequency of a message is very important for real- time
control. The frequency required is determined by vehicle dynamics and desired control
system performance.
􀀁 Priority: To ensure the timely receipt of the message, different priorities should be
assigned to different messages. The principle is that messages related to the safety and
with stringent timing requirements should have higher priority. But careful design must
also ensure that the highest priority messages do not use up too much of the available
data bus bandwidth with frequent updates and starve the delivery of other important
messages.
􀀁 Message encoding and length: To ensure that the data exchanged among functional
blocks has enough precision within its possible range, yet does not use any more of the
communication bandwidth than necessary, numerical encodings such as fixed point
limited range or integer case encoding of finite possibilities can be used. Short messages
are preferred to avoid tying up the network in the case of other urgent communication.
Error detection and correction coding is another way to ensure reliable message
transmission.
4.4 Vehicle and Lane Position Sensing
How to determine the vehicle’s lateral deviation relative to the lane center with high accuracy,
high bandwidth and robustness is very important to the successful implementation of electronic
guidance/ assist systems. Figure 4.3 shows a general schematic of vehicle and lane positioning
43
sensing. The sensing device ( e. g. GPS receiver, video camera or magnetometers) detects the
changes ( e. g. electro- magnetic wave, light or magnetic field) in the sensed infrastructure. The
position between the vehicle and the lane is then resolved by local information processing of the
sensor outputs and the result is sent to other functional blocks. Complementary sensors are
needed for some technologies to ensure robustness and accuracy. For example, an INS sensor
package is usually installed as a complementary sensor to a GPS system to mitigate blockage
situations. Different sensing technologies can also be used to complement each other in order to
increase system reliability, or to allow the system to operate in environments where different
infrastructure sensing support may be available on different parts of a route.
Figure 4.3 Schematic for Vehicle Position Sensing
4.4.1 Performance Requirements
4.4.1.1 Position accuracy:
The position accuracy depends on specific system configurations and application scenarios. For
the guidance application, the indicated position error shall be sufficient to provide both guidance
and guidance feedback to the driver in diverse operating conditions ( e. g. different weather,
visibility, signal blocking etc). As a rule of thumb, the sensed position accuracy should be
smaller than 1/ 4 to 1/ 2 of the needed tracking accuracy.
4.4.1.2 Spatial coverage:
The spatial coverage shall cover the width of the desired operating roadway.
4.4.1.3 Update rate:
The timing and update rate of sensors and signal processing shall be sufficient for achieving the
performance requirements. As a general rule, a 10 Hz update rate should be sufficient for most
bus operations, since the bus fundamental dynamics generally operate below 2 Hz.
4.4.1.4 Time delay:
Although it is always preferable to have a sensing delay as short as possible, the sensing delay
from input to output shall at worst be kept shorter than 0.05 seconds to allow accurate tracking of
bus dynamics at the 10 Hz update rate.
44
4.4.1.5 Robustness to environmental factors:
The measurements of the vehicle position sensing system shall be consistent regardless of
changes in environmental factors ( e. g. heavy rain, standing water, snow, dirt and extreme
temperature variations etc), or such factors shall be compensated.
4.4.1.6 Sensor Redundancy:
Vehicle and lane position sensing is the critical sensor for the operation of the VAA- PD and
VAA- VG systems. Depending on the exact operating scenarios and system design requirements,
redundant sensors may be necessary to ensure safe operation. The redundant sensor could be
another sensor based on the same technology ( e. g., two magnetometer bars installed at different
locations under the bus) or