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 I: Technology Assessment
UCB- ITS- PRR- 2007- 21
California PATH Research Report
Steven E. Shladover, et al.
CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS
v
Lane Assist Systems for Bus Rapid
Transit, Volume I: Technology
Assessment
Report on Technical Visit to Europe on Electronic Guidance
Technologies
Steven E. Shladover, Wei- Bin Zhang, Doug Jamison, Graham Carey, Stefano Viggiano
David Angelillo, Jim Cunradi, and Brian Sheehan, Dave Schumacher, Maurilio Oropeza,
Matthew Hardy, Walter Kulyk and Yehuda Gross
Prepared by:
University of California at Berkeley
PATH Program
1357 South 46 th Street
Richmond, CA 94804
Prepared for:
California Department of Transportation
U. S. Department of Transportation
Federal Transit Administration
Final Report for RTA 65A0160
v
v
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 direction of Walt Kulyk, Brian Cronin, and
Venkat. Pindiprolu of the Federal Transit Administration and Yehuda Gross of 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 who participated in the technical visits. Specifically, Jim Cunradi and David
Angelillo of Alameda Contra Costa County Transit ( AC Transit), Doug Jamison of
Central Florida Regional Transportation Authority ( LYNX), Graham Carey and Stefano
Viggiano of Lane Transit District ( LTD), Dave Schumacher, Maurilio ( Mario) Oropeza
and Brian Sheehan of San Diego Association of Governments ( Sandag) 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 organizations and cities who hosted our visits:
• France’s Ministry of Equipment, Transportation, Housing, Tourism and the Sea
• Rouen Regional Government ( Agglomération de Rouen)
• Connex
• Siemens/ MATRA
• Samenwerkingsverband Regio Eindhoven ( SRE) ( Regional government for
Eindhoven, Netherlands)
• Advanced Public Transport Systems ( APTS)
• TNO Automotive
• EVAG - public transit agency for Essen, Germany
• Connexxion
• ANT Consultants
FROG Navigation Systems
v
v
Abstract
This report documents the information collected by an FTA- led delegation to several
European organizations that have had experience in the development and operation of
transit lane assist systems based on three different technologies, including ( 1) optical
guidance in Rouen, France, ( 2) magnetic guidance in Eindhoven, Netherlands and ( 3)
mechanical guidance in Essen, Germany. It includes summaries of the briefings prepared
by the European hosts in response to questions from the delegation, the discussions the
delegation had with their hosts and observations based on riding the systems in public
service. This report is accompanied by a CD ROM that includes the presentations given
by the various hosts during the visit and pictures taken by the members of the delegation.
Keywords: Vehicle Highway Automation, Lane assist, electronic guidance, Bus Rapid
Transit
v
v
Table of Contents
Acknowledgements ......................................................................................................... v
1. French DOT’s Perspectives and Support on Electronic Guided Bus Rapid Transit ....... 1
1.1 Background ........................................................................................................... 1
1.2 Bus Rapid Transit ( BRT) in France........................................................................ 2
1.3 Government Support for the Development of Electronic Guidance System............ 3
1.4 “ On Tires” Light Rails, Economic Point of View................................................... 4
2. Optically Guided Bus in Rouen, France....................................................................... 6
2.1 Background .......................................................................................................... 6
2.2 Application............................................................................................................ 6
2.3 Experience............................................................................................................. 7
2.4 Observations........................................................................................................ 10
3. Phileas: Magnetic Guidance Based Advanced BRT Vehicle ...................................... 11
3.1 Background ......................................................................................................... 11
3.2 Application.......................................................................................................... 12
3.3 Experience........................................................................................................... 14
3.4 Observations........................................................................................................ 16
4. Safety Assessment and Certification.......................................................................... 18
5. Mechanically Guided Transit Buses in Essen............................................................ 20
5.1 Background ......................................................................................................... 20
5.2 Application......................................................................................................... 21
5.3 Experience.......................................................................................................... 21
5.3.1 Guideway infrastructure................................................................................ 22
5.3.2 Vehicle modifications ................................................................................... 23
5.3.3 Operational and safety issues ........................................................................ 23
5.3.4. Maintenance................................................................................................. 25
5.4 Observations....................................................................................................... 25
6. Magnetic Guided BRT and People Movers ............................................................... 26
6.1 Background ........................................................................................................ 26
6.2 Applications ....................................................................................................... 26
6.3 Experience........................................................................................................... 27
6.3.1 General guidance experience........................................................................ 27
6.3.2 Guided bus systems...................................................................................... 29
6.3.3 Automated people mover systems ................................................................. 30
6.4 Observations....................................................................................................... 31
7. Transit Operator’s View on Guided Buses ................................................................. 32
7.1 Background ......................................................................................................... 32
7.2 Applications ....................................................................................................... 32
7.3 Experience.......................................................................................................... 33
7.4 Observations....................................................................................................... 35
Appendix A................................................................................................................... 36
Final Detailed Itinerary.................................................................................................. 36
Appendix B. Europe Technical Visit Participants .......................................................... 39
Appendix C Questions Prepared Prior to the Visit ......................................................... 41
v
v
List of Figures
Figure 1. Optically Guided TEOR in Rouen……………………………………………... 8
Figure 2. TEOR Precision Docking at Bus Stop…………………………………………. 9
Figure 3. Phileas Bus in Eindhoven……………………………………………………... 13
Figure 4. Phileas Bus Docks at a Station………………………………………………... 15
Figure 5 Automated Guided Bus at Essen………………………………………….…… 22
Figure 6 A Close Look at Guiderail……………………………………………….……. 22
Figure 7 Guide Wheel………………………………………………………….………. 23
Figure 8. Automated Park Shuttle People Mover………………………………….…… 28
v
1
Lane Assist Systems in Europe
Report on Technical Visit to Europe on Transit Lane Assist
Technologies
A delegation from the U. S. spent one week visiting European organizations that have had
experience in the development and operation of transit lane assist systems. We came
with a list of questions designed to help identify the most important similarities and
differences between the needs of European and American transit operators for use of such
systems, and to help us gain the benefit of their prior operational experience. The
locations that we visited were using three different technologies for lane referencing,
providing important contrasts and points of comparison, and they also had significantly
different periods of operational experience.
The U. S. delegation members represented FTA and FHWA, as well as four different
transit properties, plus the project staff from the U. C. Berkeley PATH Program and
Mitretek. The complete list of delegates is attached as Appendix B, while the agenda for
the visit is attached as Appendix A and the list of questions that were asked at the visits is
Appendix C. This report is accompanied by a CD ROM that includes presentations given
by various hosts during the visit and pictures taken by the members of the delegation.
The primary information derived from these visits is summarized below, in the sequence
in which they were scheduled.
1. French DOT’s Perspectives and Support on Electronic Guided Bus Rapid Transit
French Ministry of Equipment, Transportation, Housing, Tourism and the Sea ( French
DOT) has been a champion of the development and deployment transit guidance system.
The delegation first met with the French DOT in Paris to learn the national perspective on
application of electronic guidance technologies to Bus Rapid Transit.
1.1 Background
The French Ministry of Equipment, Transportation, Housing, Tourism and the Sea
( French DOT) has the primary responsibility for development of new transportation
technologies. Under a large research program Prédit, the French DOT committed to
investigate innovative ideas for improvements of land transportation systems. The
program focuses on both technologies and the organizational issues involved in the
deployment and operation of advanced technologies. Prédit 1, which studied new
technology solutions to transportation issues, was carried out between 1990- 1994. The
research included subjects related to high speed rail ( TGV), automated subways, safe
cars, and environmentally friendly vehicles. Prédit 2 ( 1996- 2000) investigated
technologies for goods movement and a wide range of technologies for urban transport
options, including guidance, automated transport, car sharing, and micro- cars.
2
The Prédit 3 program ( 2002- 2006) investigates new approaches to mobility behaviors,
financing transport systems in light of recent national funding cutbacks, and provision of
multi- modal information. This program has a budget of 􀀁 305 million contributed by
various public partners involved in Prédit. The projects are grouped into the following
eleven areas:
Mobility, Territories
Group 1 – Mobility, territories and sustained development
Group 2 – Services and mobility
Security and Safety
Group 3 – New knowledge for security
Group 4 – Security oriented technologies
Goods Transportation
Group 5 – Logistics and transport of goods
Group 6 – Technologies for goods transportation
Energy, Environment
Group 7 – Energy and environmental impacts
Group 8 – Clean and energy- saving vehicles
Technology Integration
Group 9 – Integration of information and communication systems
Group 10 – Vehicles and infrastructures: integrated developments
Integrating Policies
Group 11 – Transportation Policy
Prédit 3 is organized to provide a nationwide perspective on the policies associated with
the development and deployment of new public transportation systems and technologies.
1.2 Bus Rapid Transit ( BRT) in France
Increasing the mode shift to public transportation is difficult, even in France, because in
many cases public transport does not meet the needs of the people in the same manner as
the personal vehicle. This can be especially so with bus transportation due to circuitous
routing, resulting in indirect trips to an individual’s destination. Bus Rapid Transit
systems are seen as a means of overcoming this problem by providing a more economical
mode than Light Rail Transit ( LRT). Even including the incremental cost of the
technology for guidance and exclusive lanes, the guided BRT system still costs much less
than an LRT system.
Prior to 2002, the French Government financed public transportation infrastructure,
providing 25- 30% of the capital costs for bus system lanes, shelters, bus stop pads, and
other infrastructure. Rolling stock was the responsibility of the local agencies. A new
government structure was introduced in 2002, with a move to de- centralization. This
resulted in the elimination of national assistance and shifted the responsibility entirely to
the local governments. National taxes were reduced due to the reduction in funding
provided, but local taxes have risen to make up the difference. This funding shift has
3
caused local authorities to become more focused on efficiently spending their limited
resources on transit improvements.
BRT systems were proposed, based on assumptions that the vehicles would cost the same
as traditional transit buses. Guidance is considered for precision docking and reduction
of infrastructure costs, buses operate in exclusive lanes, and the level of service is
comparable to LRT systems. Traditional bus systems cost two million euro per kilometer
( 􀀁 2 million/ km) for a bus and the associated infrastructure improvements, so the goal was
set for a BRT system to cost no more than seven million euro per kilometer ( 􀀁 7
million/ km) including vehicles, stations, lanes, and guidance. BRT operations on
highway shoulder lanes were also considered, but have not been deployed to date.
1.3 Government Support for the Development of Electronic Guidance System
Guided bus systems were developed under the Prédit 2 program. The main motivation
for use of lane assist technology in France right now is a recent law requiring
handicapped accessibility of buses within ten years, which in turn is generating demand
for precision docking. Equally important, the government has set a goal of using bus
lateral guidance to allow Bus Rapid Transit to provide LRT- like quality of service at
lower cost ($ 7 M vs. $ 15 M per km for LRT). However, the Ministry representatives
noted that LRT is currently “ fashionable” in France despite higher costs, so buses are
generally not supplanting LRT initiatives. This is particularly notable since funding for
capital costs of transit systems is now completely decentralized in France, with no
subsidies from the national government.
The selection of computer vision technology for bus guidance was based on the ability of
a domestic supplier ( MATRA, since acquired by Siemens) to provide it, without any
formal evaluation relative to other alternatives. Snow was not considered to be a
significant problem because it is quite rare in French cities ( typically only a light snowfall
a few times per year). The “ worst case” tolerance at speed including turns was
considered to be five centimeters ( 5 cm). A safety margin of forty to fifty centimeters
( 40- 50 cm) is built into the systems to allow for system failure recovery.
Certification of automatic guidance systems in France requires satisfying the codes for
both highway vehicles and for guided vehicles ( such as automated people movers).
The increased cost- consciousness of local transit authorities will shift the focus of BRT
guidance technology deployments from new vehicles to retrofits of existing fleets.
Systems will need to move from focusing on “ fashionable” new vehicles to continuing to
use their current rolling stock. The government’s goal is to have guidance systems
retrofitted on exiting rolling stock for three thousand Euros ( 􀀁 3 K) per bus, which is
much less than the cost of customized new vehicles, but it is not clear whether that is a
realistic or achievable goal. Research is also being conducted to add improvements to the
currently available guidance system, so that it could potentially be extended beyond the
current precision docking application in Rouen to use along the entire bus route by 2007.
4
The Ministry sees a serious need for outreach to local government officials, to educate
them about the opportunity to use precision docking to improve accessibility of their
buses. Additional benefits, such as reductions in stopped vehicle time, narrowing of lane
widths, and driver assistance are considered above- and- beyond this accessibility goal.
Although only one current project in France is providing lane guidance on buses between
stations ( for a limited distance in Clermont- Ferrand), future projects are expected to
extend capabilities in that direction.
1.4 “ On Tires” Light Rails, Economic Point of View
Public transport investments in France have tended in the past toward modes that were
considered “ more fashionable” than other modes. This trend led to the shift from
streetcar transport to bus transport, with the removal of much of the streetcar track and
infrastructure, similar to what was seen in many United States cities. The trend then
shifted from bus to light rail, with heavy investment in exclusive right- of- way, rails,
stations, tunnels, and other related infrastructure. The rail transport systems are more
expensive and do not provide greater benefits than buses, according to a Ministry
representative, however they have been the mode of choice and have had greater public
acceptance.
The TEOR service in Rouen, France is good example of a BRT system providing
equivalent service to LRT but at a great reduction in cost. Both LRT and BRT systems
have been deployed in Rouen, allowing a direct side- by- side comparison of capital
investment and operating costs.
The LRT system was the first to be installed in 2001. Rouen sought to expand the system
and accepted bids for this extension but found the cost to exceed available funding
resources. The city then decided to construct the new lines using BRT.
The costs of right- of- way, construction of the guideway, stations, and ticket- vending
facilities are the same for each mode. Rail, however, has added costs associated with
rails, customized garage and maintenance facilities, energy distribution system, rolling
stock, and engineering. Both systems were built under bids that included a complete
reconstruction of the associated corridor including surrounding vehicle travel lanes,
stations, signals, and vehicles, so specific costs associated with each element cannot be
provided. It was found, however, that the cost per kilometer of the BRT was forty
percent ( 40%) less than that of LRT. The cost of overhead catenary wires to provide
electrical power to LRT vehicles in the corridor served by TEOR was equivalent to
eighty ( 80) articulated bus vehicles.
A side- by- side comparison of bus and LRT costs in Rouen ( which has both types of
systems) showed operating cost advantages for buses, based on lower maintenance for
both vehicles and right of way, and even larger cost advantages when the capital costs
were annualized, as shown in the table below:
5
Annual Operating
Costs
Total Annualized Costs
( Includes Amortized Capital and
Operating Costs)
Non- guided Buses 􀀁 6,700,000 􀀁 18,000,000
BRT TEOR Guided Buses 􀀁 6,750,000 􀀁 24,000,000
LRT 􀀁 8,000,000 􀀁 32,000,000
It is noted that driver costs for BRT are more than LRT, however overall operational
costs for BRT are less than LRT. Maintenance cost for LRT is estimated to be four to
five ( 4- 5) times the cost of BRT, mostly due to the limited number of vendors and
customization of parts. The City of Rouen considers annualized costs of their
investments, which include operating costs plus capital investment amortization at eight
percent ( 8%) over sixty ( 60) years.
Financing issues will still need to be addressed for deployment of future systems and
expansion of existing ones. The recent shift from national to local funding has resulted in
local authorities placing more emphasis on economical solutions while the riding public
is used to more “ luxury” systems that were subsidized by the national government’s
capital cost funding. Around sixty to seventy percent ( 60%- 70%) of the riders are youth
and students, so using farebox recovery to pay for “ luxury” infrastructure puts too great a
burden on a typically lower income rider.
6
2. Optically Guided Bus in Rouen, France
Optical guidance system has been deployed in Rouen, France for a number of years. The
delegation visited city of Rouen and met with Rouen Regional Government
( Agglomération de Rouen), the transit operating company Connex and the technology
developer Siemens/ MATRA.
2.1 Background
MATRA, the French aerospace company ( which also has experience in automated
guideway transit and railway control systems), developed the optical guidance system for
the CiViS bus, which is now also being used on another bus, the Agora model from
Irisbus. MATRA was subsequently acquired by Siemens. The optical guidance system
is currently being used for precision docking in Rouen, but in Clermont- Ferrand it is also
being used for lane guidance between stations on a relatively short section of track ( about
1 km). Rouen has been a pioneer in deployment of the Siemens/ MATRA precision
docking on its TEOR bus services, representing three east- west bus lines with high
passenger volumes.
The Rouen urban area has a population of about 400 K, and the local bus system serves
163 K passengers per day. Of this total, 30 K passenger trips are on the TEOR lines, the
BRT services which have precision docking. The overall farebox recovery ratio for the
system is about 40%. The main efficiency measure used by the transit authority is the
number of passenger trips per vehicle kilometer. The values of this measure for the three
primary transit modes are 10.67 for the LRT system ( part of which runs as an
underground Metro service), 4.13 for the TEOR buses and 1.76 for the remaining buses.
The capital costs of the system are funded by multiple layers of government agencies, but
the operations are the responsibility of a private concession operator, Connex, who are
paid based on their ability to provide the quality of transit service specified by the public
sector sponsors. This visit included a briefing and rides on the optically- guided buses, as
well as rides on the LRT system that provides north- south trunk transit service in Rouen.
2.2 Application
The TEOR BRT service was initiated after the costs for a new LRT service on the
intended route became too high ( twice the amount budgeted). It was still initiated as a
comprehensive project analogous to LRT, involving the complete curb- to- curb
reconstruction of the streets on which it runs. This provided for one or two bus- only
lanes in most places, with a distinctive pavement color based on use of red aggregate in
order to discourage ‘ normal’ drivers from trying to drive in the bus lane.
The system planners believed that it was important to provide both enhanced image and
improved quality of service ( trip time and reliability) in order to attract new choice riders.
The en- route trip time and reliability improvements are mainly associated with use of
traffic signal pre- emption ( at most intersections, except when crossing national roads)
7
and with the use of one or two dedicated bus- only lanes. These are estimated to save 6%
of the overall travel time, while the precision docking is estimated to save another 4% of
travel time, but it is not clear how much supporting data are available to back up these
estimates.
The system was designed to emulate LRT as much as practicable, including similar
station designs and spacings ( 500 m) and even the use of a similar- sounding bell- like
alert to warn passengers about imminent door closure. They stop and open doors at all
stations and operate on similar headways to LRT.
2.3 Experience
System costs are generally estimated per route kilometer, incorporating the costs of both
vehicles and infrastructure. They estimate that their guidance system adds a cost of about
􀀁 500K ($ 650 K) per kilometer, which is less than 10% of the total cost of 􀀁 6.3 M ($ 8.2
M). By contrast, an LRT system would cost 􀀁 15 M to 􀀁 20 M per km ($ 20 to 26 M) in the
same environment.
Since the start of precision docking service in 2001, the TEOR system has performed a
total of over 5 million docking maneuvers. Although no formal study has been made of
the dwell time savings from precision docking, the MATRA representative offered an
off- the- cuff estimate that former dwell times of 40 seconds were being reduced to the
range of 15- 20 seconds in Rouen. Although there have been some failures, none of them
have led to movement of the bus outside the prescribed “ safety envelope”.
The speed limit for precision docking approaches ( typically, 40 km/ h) is painted on the
pavement upstream of each station to remind the drivers of this limit, since the drivers are
completely responsible for controlling bus speed. The speed limit marking was required
as part of the certification process, but the system owners and operators do not see any
need for automatic speed control of the buses to limit the docking speed.
Rouen chose to replace the originally- planned CiViS buses with more conventional
Agora model buses after testing the first two CiViS buses. They had problems primarily
with the CiViS’ hybrid propulsion system, particularly regarding poor fuel economy and
reliability and high acoustic noise levels.
At some stations, the guidance line follows a complicated trajectory, with built- in
overshoot. The design of the layout of the guidance line requires some sophistication,
including use of a computer simulation of vehicle response, and is a significant cost item
associated with system implementation. It starts upstream of the stop, and needs at least
20 m of straight line prior to the stop in order to allow the articulated bus to straighten out
so that all door openings are close to the platform.
When drivers start to work on the TEOR lines, they have a 5- day training period to adjust
to the new vehicle and guidance system, as well as to learn how to activate the signal pre-emption
system. Both the TEOR drivers and the LRT drivers are required to drive a
8
regular bus route for one week of each 12- week schedule period in order to ensure that
they maintain their driving, and especially their docking, skills.
Figure 1. Optically Guided TEOR in Rouen
Surveys of passengers have shown comparable levels of satisfaction among passengers
on the TEOR bus lines and on the LRT system, which was one of the primary goals for
the system. Surveys of the bus drivers have also found them generally favorable to use of
the guidance system, especially in the locations where the dedicated lanes are available to
segregate them from other traffic. They find it more convenient than doing the docking
themselves and it also relieves them of some workload so that they can interact more with
the passengers. Bus drivers also say that it helps reduce their level of stress. They
believe that it improves their vigilance by letting them devote more attention to their
surroundings, rather than low- level vehicle maneuvering. As drivers have become more
comfortable with the guidance system, they have even begun to ask for extensions to the
guidance lines to help them exit stations with complicated geometry or with difficult
maneuver profiles. Based on limited questioning of drivers during our visit, they appear
to like the lane assist system as it is now for docking, but do not seem to be interested in
seeing it extended to use between stations. However, Siemens plans to provide
continuous lane guidance between stations by 2007 in Rouen and Bologna. Siemens
believes that the distinctive paint stripe markings that serve as their guidance reference
can be continued through intersections without necessarily confusing drivers because
9
they are no more visually distracting than trolley tracks, which are frequently
encountered through intersections.
Even though the precision
docking helps speed up
passenger boarding and
alighting, the full potential of
this cannot be gained because of
the need for passengers to queue
up at the ticket validation
machines near the doors.
Political difficulties have
impeded the development of a
new national fare collection
system in France, which could
ultimately reduce the bottlenecks
associated with fare collection
and validation
Figure 2. TEOR Precision Docking at Bus Stop
Siemens said that the vision based guidance system does better than the driver at
detecting the guidance lines under bad lighting conditions because of the low mounting
height of the camera and the use of sophisticated filtering, but they noted that it still has a
few problems. The system sometimes can not engage when the sun angle is low and
pointed toward the bus, as well as in the special condition of glare off wet pavement. The
drivers do not have any indicators of how far along the platform to stop the bus, which
led to some of the drivers over- shooting the platform on our demonstration rides and
stopping so far forward that the platform edge was already receding from the curb line,
leaving a larger than intended gap for passengers at the front door. There are only a few
snowy days per year in Rouen, and drivers are advised to turn off the system when it is
snowing. The guidance system occasionally experiences failures. When it fails, the
driver is informed and will take over and control the bus. The transit operating company
conducts fault diagnoses and replaces components under most circumstances.
French cities use asphalt pavement rather than concrete, regardless of the density and
weight of the traffic on that pavement. Along the TEOR route, there was noticeable
pavement rutting at several stations, after 4 years of service. They are not concerned
about this, but said that they expected to replace the top 5 cm layer of pavement after 7 to
10 years. At one station, the city said that the paving was not done correctly and they are
pursuing the contractor to replace it. The painted reference stripes on the road surface
need to be cleaned off periodically and then repainted annually, according to the public
agency people ( although Siemens/ MATRA refers to it as annual “ refreshing”). They use
a template to improve the accuracy of the painting.
The Siemens/ MATRA guidance system is provided with a 2- year warranty, after which
replacement parts are offered through a service contract. The system is expected to
10
continue in service for ten years before needing a major overhaul, so that one major
overhaul would be contemplated within the 20 – year typical operating lifetime of a bus.
The on- site regular maintenance work is done by the maintenance staff of the transit
operating company.
2.4 Observations
Rouen has successfully implemented the optical guidance system for precision docking,
while choosing not to apply it for the broader capability of automatic steering between
stations. They are expanding the TEOR services, with the major investments associated
with reconstructing the streets for smoother running surfaces and separated bus lanes
where there is sufficient room. In their TEOR system as a whole, the major service
improvements appear to be in the provision of newly constructed bus- only lanes and the
traffic signal pre- emption, rather than the precision docking.
11
3. Phileas: Magnetic Guidance Based Advanced BRT Vehicle
The city of Eindhoven in the Netherlands is developing and deploying an Advanced Bus
Rapid Transit System, Phileas, in collaboration of Advanced Public Transport Systems
( APTS) and Frog Navigation Systems.
3.1 Background
Samenwerkingsverband Regio Eindhoven ( SRE) is a regional government body in the
Eindhoven region comprising 21 municipalities with a population of about 700,000, and
has the lead responsibility for deployment of the Phileas guided bus -- ‘ tram on tires’ or
advanced Bus Rapid Transit. This is being done in cooperation with their local transit
operations contractor, Hermes, and the developer of the bus, Advanced Public Transport
Systems ( APTS- a subsidiary corporation of VDL Group BV), a local company. A large
group of public agencies and a few private entities contributed resources totaling 􀀁 115 M
( about $ 150 M) to development of a demonstration project to serve two goals -- to
improve local public transportation and to provide a stimulus to the local vehicle industry
– by testing a guided bus system. The major contributors included:
- National government – 􀀁 48 M
- Province – 􀀁 2 M
- Municipalities – 􀀁 45 M ( mainly for infrastructure)
- European Union economic stimulus program – 􀀁 9 M
- Transportcom ( bus company) – 􀀁 11 M
These costs included the design and construction of a 15 km concrete track in the median
of existing roads ( in the process reconstructing those roadways from curb to curb), the
development of a completely new bus, starting from a clean sheet of paper, and the
acquisition of twelve of those buses. The system was designed to be the functional
equivalent of an LRT system, but using rubber tire vehicle technology. It was
implemented as part of a broader local transportation policy that also includes increasing
the cost of parking tickets and encouraging bicycle usage for access to and from stations.
The project began in 1993 with the planning for the western/ new growth district of
Eindhoven and the official start of the demonstration project commenced in 1999 with an
order for 12 Phileas vehicles. The public ( customers and bus drivers) was involved
( through focus groups) on the selection of the exterior and interior design. SRE utilized
an economic development framework for creating the Phileas BRT system. This
comprehensive approach included a completely new vehicle design, re- configured
infrastructure and ROW and a public policy and logistics component. The framework for
the BRT system and the comprehensive scope of the project dictated the decision to test
lane assist/ electronic guidance technology.
The primary transportation goals of the Phileas project were to provide circulation with
regularity and punctuality. Secondary goals included improved service frequency, speed,
reliability, flexibility, and comfort, with an ecologically sound and a modern design. It is
unclear if the lane assistance technology was introduced to accomplish a primary or
12
secondary goal of the project. However, the lane assistance technology is an integral
component of the intelligent vehicle concept that allows for the vehicle location, speed
and stops to be tracked. The expectation is that the guidance features would enhance safe
operation of the vehicle as well.
The buses were gradually introduced to public service last summer, without the guidance
function. Up to seven vehicles were operated for up to 16 hours per day, and in October
the guidance function was implemented with considerable public fanfare. By the end of
November, all of the Phileas buses were removed from public service for upgrades, based
on a variety of reports of needed improvements. They are expected to be gradually re-introduced
to service starting in February, but the guidance function will not be reinstated
until at least another three or four months. Our group was fortunate to get a special
demonstration ride on one of the buses during our visit, despite their having been
withdrawn from public service.
3.2 Application
The eventual application is intended to be a network of busways serving the Eindhoven
region, but the initial line is 15 km long, connecting the Eindhoven airport to the central
downtown railway station neighborhood, passing through a variety of residential and
industrial neighborhoods. The Phileas bus was designed to provide this service, but also
to help revitalize industry in the region. The design of the bus is unusual in a variety of
aspects, including not only the lane assist system, but also:
- Alstom series hybrid propulsion system, with electric drive motors at each wheel
except for the front axle, and regenerative braking ( in the future it could be replaced
by the Allison parallel hybrid, with only the rear- most axle being powered)
- battery energy storage sufficient to propel the bus for 3 km without the engine
- independent active steering of all axles
- completely low floor
- lightweight, modular composite body structure constructed by Fokker, with 18 m and
24 m versions ( total body weight for 24 m version is only 4700 kg)
- automatic speed control
- Interior image – doors on both sides, adaptable, suspended furniture
- customizable exterior shape ( initial design chosen by the Eindhoven region).
The combination of the lane assist system with other innovative features makes this a
particularly complicated system to study.
The Phileas electronic guidance system is based on an intelligent vehicle that ‘ knows
where it is’ using magnetic markers embedded in the road surface every 4- 5 m ( 13- 16
feet). The magnetic guidance allows the vehicles to operate at maximum allowable
speeds within the city ( 50 kph) and up to 70 kph under most weather conditions.
Although lateral guidance was designed throughout the corridor, the dimensions of the
vehicle travel lane were only narrowed modestly from the standard width to allow for
13
standard buses to share the dedicated ROW. The electronic guidance is designed to allow
for precise station docking.
Figure 3. Phileas Bus in Eindhoven
The guidance system is applied within the most ambitious operational concept of any of
the systems we visited, involving use of both semi- automated ( automatic steering) and
fully automated ( automatic steering and speed control) modes. Automated mode allows
for full electronic control of the vehicle’s braking, throttle and steering mechanisms.
Semi- automatic mode controls the Phileas’ steering while the driver controls the throttle
and brake. These automated modes were applied not only in the dedicated bus lanes in
the medians of the roadways, but also in mixed traffic while negotiating rotaries between
sections of dedicated bus lanes.
The development of the Phileas bus represented an investment of about 􀀁 40 M ($ 52 M),
of which 􀀁 4 to 􀀁 5 M ($ 5.2 to $ 6.5 M) was associated with the adaptation of the FROG
magnetic guidance system for use on the bus. The initial purchase of buses for use in
Eindhoven was for twelve of the 18 m articulated buses. The cost of the buses was not
broken out explicitly from the costs of the complete demonstration project, but APTS
quoted approximate purchase prices of 􀀁 1.1 M and 􀀁 1.34 M ($ 1.43 M and $ 1.74 M) for
the basic 18 m and 24 m buses respectively.
14
The vehicle has undergone numerous tests, including the guidance system at speeds up to
75 km/ hour. The claim was made that magnetic guidance ultimately will be more
reliable and cheaper to implement than optical guidance, although no specific cost figures
were available.
3.3 Experience
A distinct characteristic of the Phileas system is that it was developed locally and entirely
from the ground- up. Although costly, this approach has merit in the ability to effectively
brand a new type of modern transit service that is competitive with light- rail and/ or
streetcars in terms of comfort, performance and costs. The hybrid propulsion system on
the bus is expected to save about 30% energy consumption compared to a conventional
powertrain. The ride quality is equal to and in some ways superior to that of light-rail/
streetcar vehicles. In terms of noise, the Phileas claims to operate more quietly than a
streetcar ( although no noise studies have been done to prove this yet). However, starting
and stopping still show the vehicle performance to be less smooth than a rail car. It
should be said that current performance in this category is noticeably better than
competitors and once the electronic guidance technology is more fully developed, the
Phileas could match rail technologies in this area.
The electronic guidance system was chosen in order to provide precision docking at
stations and enhance ride comfort between stations. It also enabled a modest reduction in
lane widths on the busway, from the normal 7.2 m width for a pair of bus lanes to 6.6 m
here ( accommodating a bus that is 2.55 m wide). However, that busway is currently
being used by conventional buses without lateral guidance providing normal public
service, so clearly they can also be driven within that width of lane. The guidance system
is not designed to directly follow the path defined by the sequence of buried neodymium
magnets, which are 30 mm long, 15 mm in diameter and installed 4 to 5 m apart. Rather,
the path is defined by multiple manually- steered drives along the track to identify where
the drivers drive, and that trajectory is then smoothed out to reduce jerk. According to
the transit operator, this trajectory smoothing is not yet perfected and still requires
another 3 to 4 months of adjustments before the buses can be returned to automatic
service. There is also one section of track with a large amount of reinforcing steel that
distorts the magnetic field of the guidance magnets, leading to some ride roughness ( both
lateral and longitudinal).
In the automatic mode, the bus follows a pre- defined speed trajectory to stop at each
station. When the bus operates in the automatic or semi- automatic mode, all stops are
programmed as in a tram, and the driver cannot skip stations. The bus does not have any
buttons for passengers to request a stop, nor does it have door opening buttons, because
the doors always open automatically at the station. When the road conditions are
slippery, the semi- automatic mode can be used to provide automatic steering, but the
driver must then control the speed.
The Phileas bus uses an all wheel steering design with each axle of the bus being steered
independently, using a triple- redundant computer system, based on measurements of
15
lateral displacement at that axle, in order to emulate LRT tram steering along a track.
This independent steering makes it possible for the bus to do a “ crab steering” lateral
motion to dock at each bus station, requiring less approach distance than a conventional
bus to line up all doors parallel with the platform. The all- wheel steering also provides
high maneuverability, to the extent that the bus developers claim that it drives like a car
rather than a bus, and that the drivers do not take full advantage of its handling
characteristics until they learn it well. On the approach to the Eindhoven Airport, it
successfully negotiated a very tight turn. The developers cite a turning radius of 11.8 m
to the outer side of the bus.
The defined positioning tolerances are 1 cm at stations and 5 cm at a speed of 50 km/ h.
The gap at docking stations was reported to be 6 cm, but when we rode the bus it
appeared to be more like 10 cm, which is larger than shown on videos of vehicle tests.
The larger gap appears to represent an additional safety margin, but once the system is
proven reliable, the nominal gap could be reduced to the 3 cm range.
Figure 4. Phileas Bus Docks at a Station
The electronic guidance system can only be used at speeds up to 45 km/ h on the public
route in Eindhoven, but the bus can be driven manually up to 85 km/ h. When the bus
enters the guideway equipped with magnets, it needs to read about 100 m of magnets in
order to recognize its location and activate the guidance function. If the bus deviates
16
from its intended path by more than the 25 cm maximum tolerance, it will be
automatically shut down ( described as “ electronic wall”).
Traffic signal pre- emption ( emulating tram operations) and automatic vehicle speed
control are used to maintain schedule times. If the bus gets ahead of schedule, the speed
control system can slow it down accordingly, and the speed control appears to be
integrated with the signal pre- emption system, although the specifics of how that works
were not discussed.
The Driver Vehicle Interface uses a text display that identifies the current mode of
driving ( manual, semi- automatic or automatic), as well as the distance to the next station
and time ahead of or behind schedule. Kill switches to disable the automation functions
are in a console to the driver’s left. Driver training was reported to take about one day,
but there did not appear to be as formal a training program as in Rouen. No special
measures are taken to ensure driver vigilance while operating in the fully automated
driving mode. The drivers are expected to operate more like tram drivers than bus
drivers, which may require some more fundamental shifts in their culture and training.
Drivers responded most positively to the specially designed cockpit for operating the
vehicle.
The comprehensive nature of the project required significant R& D efforts and has greatly
added to the complexity of the BRT project. The complexity of the vehicle has led to a
number of issues that still are not fully resolved. These included a variety of problems
not directly related to the guidance function such as doors occasionally popping open by
about 5 cm while the bus was in motion, engine shutting down unexpectedly, loss of
hydraulic steering assist when the engine dies ( severely restricting maneuverability of the
bus), and difficulty starting the engine. Releasing the vehicle into public service before
the technology was fully operational proved to be problematic. The Phileas project has
received negative press in the region and as a result, was perhaps pressured into taking all
of the vehicles off the road simultaneously until the problems could be resolved. The
perfection of the guidance system ( responsibility of FROG) is taking longer than
expected.
The automation functions did not require any special certification for road use in the
Netherlands, where the certification authorities were comfortable assuming that the driver
remains fully responsible for the operation of the bus. The certification process is 90%
paperwork, and does not involve any testing. The manufacturer provides a 2 year
warranty on the Phileas bus, and the bus body has a 20- year warranty.
3.4 Observations
The Phileas operation in Eindhoven was the most audacious application of vehicle
guidance that our delegation encountered, in that it included fully automated driving
along most of the route, including in mixed traffic at rotaries ( where it has priority for
right of way, but not exclusive use). It was also a remarkably ambitious undertaking
from the broader perspective, in that its sponsors committed $ 150 M to a demonstration
17
project in order to determine whether it could be successful. This is equivalent to what
we would call a Field Operational Test in the U. S., rather than a full deployment of a
mature technology.
When its functions were all working, the Phileas bus displayed excellent performance.
However, it is not yet fully debugged. During the demonstration ride, there were two
occasions when the bus clearly encountered problems, the driver reported an error
message on the display and the propulsion and power steering systems were shut down.
There are questions about whether the inherent complexity of its design will make it
possible for it to be sufficiently reliable and easy to maintain for transit operators to
accept it on a broad scale.
The project also clearly demonstrates the advantage of a smooth running way in
achieving a quality of ride comparable with rail. It was unclear to some how significant
an impact the lane assist technology had on the ride quality throughout the corridor,
because in sections where the pavement had not been improved, ride quality was
negatively affected.
The cost and complexity of the Phileas vehicle is the biggest drawback to transferability
to the U. S. market. Given that the vehicle is still in a demo project status, perfecting the
outstanding problems is needed before transit agencies would likely be interested. The
potential is there, however, given the attractiveness and many desirable features the
vehicle potentially offers. However, the developers of the Phileas bus noted that they do
not have any additional firm orders for vehicles, although they are under consideration
for orders for another 10 to 15 units in other cities. If the Eindhoven network is built out
to the full extent originally planned, it would use a total of 50 buses. The bus
manufacturer needs a minimum of 7 to 10 vehicle orders per year in order to remain
viable, so there are questions about whether they will be able to survive. With the
completion of the initial Eindhoven order, they are vacating their current office and
assembly facility in Eindhoven and relocating a skeleton staff to the site of a sister
company to await the next order. They stated that in order for them to participate in the
U. S. market they would need to make some ( unspecified) adaptations to the bus and to
find a marketing partner in this country. They are also concerned about the more difficult
liability environment here. Their strategy in Europe is to market to cities that are oriented
toward trams rather than buses, because they want to be considered as a more cost-effective
rubber- tired equivalent to conventional trams rather than an expensive bus
( compared to the normal price of about 􀀁 400 K ($ 520 K) for an 18 m articulated bus in
Europe).
The magnetic guidance concept is an attractive feature that holds promise for the U. S.
market because it can be used for all weather and a variety of operating conditions.
Because Eindhoven’s design of the busway considered the dual use by both guided and
manually driven buses, the benefit of operating the bus on narrow lanes was not apparent,
though one can envision that this benefit would be substantial if the narrow lane design is
applied.
18
4. Safety Assessment and Certification
The delegation met with TNO in the Netherlands to learn about safety certification issues
related to guidance technologies.
TNO Automotive is a part of TNO, the largest research company in the Netherlands
employing a total of 5000 people. TNO Automotive has been developing a uniform
approach, called “ Integral Safety of ITS,” to apply to safety assessment and certification
for both fully automated and partially automated vehicle control systems, recognizing the
need for an approach that is simple enough to be practical to apply, but that also accounts
for the inherent complexity of automated vehicle systems. This complexity includes
challenges based on:
• control partly taken away from the driver
• complicated input/ output relationships
• tightly integrated systems
• impossible to do simple tests on systems this complex
• there are no accepted certification standards.
TNO has developed a certification approach that follows the following principles:
( a) System safety must be incorporated in the initial system design. A safety critical
system design that does not consider safety requirements at the early stage will make
it difficult to achieve safety goals.
( b) Define an accepted safety goal, based on risks that people are already willing to
accept. TNO’s philosophy on defining a safety goal for automated systems is to
design vehicle control systems as safe as necessary rather than as safe as possible in
order to make the system affordable and they believe that a completely safe system
design, even if can be achieved technically, will make the system overly complex and
unaffordable. Therefore, the safety levels of existing transportation systems have
been referenced. In Europe, the current fatality rates of existing automobile, bus and
train systems are 6.4, 6.4, and 3.0 per 1 billion traveled kilometers respectively. TNO
recommends that the fatality rate be twice as good as that of current road traffic for
new automated systems, which translates into no more than 3 fatalities per 10 9
kilometers of vehicle operation.
( c) Analyze system using standard methods such as failure mode effects and criticality
analysis ( FMECA), and identify the likelihood of occurrence of failures of various
levels of severity, focusing primarily on the failures that could cause fatalities. TNO
has defined five levels of severity and five levels of likelihood of occurrence for each,
and has put them into a matrix, in which each cell has a numerical entry showing the
acceptable level for that combination of severity and likelihood.
( d) Establish if the system meets the acceptable safety level through analysis of typical
operating conditions, considering the expected mean time between failures. It is
19
important that the evaluation enables assessment of the system as a whole, not just
components.
( e) Check compliance by means of technical tests, preferably not on public roads ( which
would require government permission).
It is recognized that some stages of the evaluation and certification process are dependent
on analyst judgment, so it is not possible to make them totally objective. However, if
multiple independent analysts go through the process in parallel, their respective
subjective biases should be cancelled out and the results could be accepted as objective.
TNO hopes that the industry establishes a ‘ code of practice’ for safety assessment and
certification for automated systems.
Systems that successfully pass the review through this approach could be issued a TNO
certificate of compliance, but it was not clear what the liability consequences would be
for TNO if a certified system produced a fatality. Current European laws require a driver
to be in control of any vehicle operating on public roads. It is also not clear what are the
liability implications if the system is under full automated control and the driver is not
given adequate response time to intervene in the operation when a failure occurs. In
private locations, laws regarding driver- in- control would not apply ( such as the driverless
people movers at Schiphol Airport, the Rivium business park near Rotterdam or the
Floriade flower exhibition).
20
5. Mechanically Guided Transit Buses in Essen
Essener Verkehrs, AG ( EVAG - public transit agency for Essen, Germany) is the first to
deploy mechanical guidance system for transit buses on dedicated busways and the
system has been in operation since 1980. The delegation visited EVAG to learn their
knowledge and experience on the mechanical guidance.
5.1 Background
Essen, Germany, is a medium- sized city in the Ruhr Valley, with a population of about
600,000 people. The city was born out of the coal- mining and steel production
industries, similar to Pittsburgh in the U. S. but today the heavy industry is no longer in
existence. Its population is aging and the community is decentralizing. Approximately
120 million passengers a year are carried by public transport in the Essen region, with
half the passengers carried by bus and half carried on rail. The transit usage of 125 trips
per year per inhabitant is in the medium range for Germany. Within the central part of
the urbanized area, all homes must be within a five- minute walk of a transit stop and for
the entire region, they must be within a ten- minute walk.
EVAG, the public transportation system in Essen, operates a network of streetcars that
has been in operation for 110 years. The city also operates 230 buses which provide bus
service to neighboring cities. In the ‘ 90s low- floor LRT was introduced. The city is
converting its streetcar lines into LRT lines as finances permit. In the heart of the city,
the LRT lines are underground. Currently there are seven streetcar lines ( total of 80
vehicles) of one- meter gauge and three LRT lines ( total of 50 trains) of 1.43- meter gauge
with high platforms.
Guided buses were introduced in the early 1980s on former streetcar lines, at a time when
streetcars were “ out of fashion”, representing a form of “ modernization” that could be
implemented within the existing right of way. Up until that point, the federal and state
governments funded 90 percent of the costs of construction of urban rail services. The
poor image of the streetcar resulted in the state adopting a policy to replace the existing
streetcar routes with heavy rail services on higher- volume lines, while lower volume lines
were being converted to bus services. Consequently, the state funding priority moved
towards heavy rail lines. With the narrow rights of way in the city, the city needed a
technology that allowed buses to be operated in a limited right- of- way. EVAG viewed a
demonstration project of new bus guidance technology as an alternative way of upgrading
the streetcar routes in order to provide improved customer acceptance and image.
EVAG researched both the wire guidance and “ O- bahn” mechanical guidance concepts
developed in the 1970s by Daimler Benz. EVAG chose to use mechanical guidance over
electronic ( wire guided, using electromagnetic induction) because of its simplicity and
the fact that it did not need an independent mechanical back- up system for safety
purposes. The electronic technology available at the time, which was field tested in
Fürth, a suburb of Nürnburg, required active electric current in a buried wire, which
21
introduced a significant maintenance challenge and a single- point failure mode. Essen
was the first operator to use guided buses in Germany and, since it was a transportation
technology research and demonstration program, 75 percent of the project funding was
from the German Ministry of Research and Technology. The construction took
approximately six months and the system began operation in September 1980. Since the
1980s, the city established a total of four guided bus routes, including the 1.4- kilometer
route Fulerumer Strasse to Wickenburgstrasse in the southwest of the city, the 0.9-
kilometer Wittenbergstrasse to Stadtwaldplatz in the south of the city, the 3.9- kilometer
Dortmunder Strasse, a median- running system on the A40 Autobahn to Kray, and a 3.0-
kilometer section running through the central tramway tunnel. In 1997, the guided bus
line that operated underground in a tunnel shared with streetcars was converted back to
streetcar- only operation.
The current buses are articulated Mercedes, between 8 and 12 years old, already
representing replacements of the original guided buses. Some of the original buses had
dual propulsion systems, with diesel propulsion for use above ground and electric trolley
propulsion for use in tunnels, but these were found to be too heavy and expensive so they
were sold off. They currently operate 15 two- axle guided buses and 31 three- axle
articulated guided buses.
EVAG does not expect any expansions to the current guided bus system because light rail
and streetcars are now back in fashion and will therefore be selected for any future
upgrades of bus lines to higher levels of service.
5.2 Application
The Essen guided buses have three pairs of guidewheels; one at the front axle, one at the
middle axle, and one on the rear section, connected to the main body of the vehicle. The
O- bahn guideway is made of prefabricated concrete track sections installed atop concrete
“ sleepers” ( like railroad ties), which are on a specially constructed foundation. The track
sections have an L- shaped cross- section, with a curb that is 18 cm high, to serve as the
running surface for a horizontally- mounted guidewheel. The width between curbs is 2.6
m for use by a 2.5 m width bus, and the guidewheels are installed to tight tolerances so
that they are slightly “ squeezed” between the curbs, maintaining contact on both sides of
the bus. The maximum speed for guided operation is 75 km/ h, but the same technology
is used in Adelaide, Australia up to 100 km/ h. The cost of the vehicle- related guidance
equipment accounts for approximately 3% of the purchase cost of a standard bus.
Construction costs for the running ways for the guided buses are comparable to those for
a streetcar. Since its operation, there has been little maintenance needed on the tracks.
Because the system uses buses, the legal requirements for light rail don’t apply.
5.3 Experience
Essen has had longer experience with guided buses than any other city, and they were
able to share some important lessons learned from that experience.
22
5.3.1 Guideway infrastructure
The construction of the busway needs to
be done to tight tolerances, and poured-in-
place concrete was judged to be
insufficient, leading to poor ride quality.
Steel guidance rails and asphalt paving
were also judged to be insufficiently
smooth. The busway geometry includes
some fairly complicated curve sections,
with overshoots beyond the expected
trajectory of the bus. The prefabricated
sections are six to eight meters long, and
are either straight or in a limited number
of constant- radius curvatures, generally
without spirals. Transitions between
sections sometimes involve steel side
rails if they do not fit one of the
standard profiles.
The main drawback of mechanical
guidance is the need for the curb, which
makes it unsuitable for pedestrian areas
and is a problem in areas with a high
number of driveways and other curb
interruptions. To accommodate
pedestrians, the guidance curbs are
Figure 5 Automated Guided Bus at Essen
designed to have gaps for pedestrian crossings up to 4 m wide. If a larger gap is needed,
such as at a street intersection, a tapered entrance section is needed for the bus to re- enter
guidance, at a speed of about 30 km/ h. No ride quality problems were observed
associated with the initiation of guidance at these locations.
The tight construction tolerances mean
that the cost of busway construction was
found to be essentially the same as the
cost of track construction for a streetcar
line ( exclusive of the electrification costs).
This means that this form of bus guidance
does not have the lower infrastructure cost
advantages typical of electronic guidance
systems. However, the guideway has
required negligible maintenance during its
25 years of operation.
Figure 6 A Close Look at Guiderail
23
Busway running surface maintenance only appears to involve repair of the edges of the
individual prefabricated elements, at the expansion joints. These repairs are done with a
concrete filler that cures over night or over a weekend when the busway is removed from
service.
The delegation members observed variability in the quality of ride when we rode the bus
on the busway in the freeway median, with some sections being noticeably rougher than
others. In the rough sections, the vibrations appeared to be associated with the joints
between prefabricated sections, which are 6 to 8 m long.
5.3.2 Vehicle modifications
The modification needed to the bus for this form of guidance is merely the attachment of
guidewheels to the axles of the bus. The front axle guidewheels are used to steer the bus,
while the guidewheels on the trailing axles are used to keep those tires from scuffing
against the curbs. These experience significantly higher forces, and therefore higher
wear, than the front guidewheels. The bus cannot be driven in reverse on the guided
track, because the guidewheels do not provide correct steering guidance and the driver
would be fighting their effects, leading to potential derailment of the bus. EVAG has not
tried bi- articulated buses on the guideway because this would require active steering of
the rear axle. The current buses use old technology axles. The last buses were ordered
for this system in 1989 with conventional axles.
The guidewheels are not expensive, but
there is currently a problem finding an
axle manufacturer willing to approve the
installation of the guidewheels without
voiding the axle warranty. EVAG does
not expect this problem to be solved until
after 2008 because until that time the bus
manufacturers are preoccupied with
developing new propulsion systems to
meet stricter emissions regulations.
Figure 7 Guide Wheel
5.3.3 Operational and safety issues
In earlier stages, EVAG experimented with in- track switches, but found them too
complicated. Therefore, the switches are now all open sections, where the driver steers to
one side or the other. There was one instance in which a driver crashed into the gore
point of the open switch. Because of the tight tolerances on the guideway and vehicles,
the docking at stations is achieved with very small gaps ( 5 cm).
24
The most common form of failure is breakage of the guidance arm if the driver hits a
curb with it at a normal bus stop. This is most likely to occur with the sweep of the rear
of the bus as it departs the stop, but still only occurs once or twice a year. The
guidewheels are designed to run at about the same height as the curb in the guideway ( 18
centimeter curb). When buses are running in normal streets, the curbs ( 13 centimeters
high) can be problematic if the driver makes contact with a curb in an unguided section.
The guide arms have a small cut which acts as a weak point for the guide wheel to break
off without damaging the entire guidance arm or the axle. A cable is attached in order to
retain the wheel in case it breaks off.
Environmental factors can also impact the system operation. A potential, but rare,
problem is the buildup of compressed snow of 3 cm or more, which could potentially
cause a bus to climb above the guide rail. EVAG removes snow from the track by
putting a snowplow on a service vehicle, which can also be used to spray salt on the
running surface.
If a bus breaks down on the busway, a service vehicle needs to fetch a towbar, which the
following bus can then use to push the failed bus. This is reported to take about 15
minutes, but such breakdowns have been infrequent.
One of the big issues in introducing the system was concern about the effects of a tire
burst on a guidewheel. To overcome this, the guidewheels are not inflatable, and a safety
wheel was installed inside the rubber coating of the guidewheel, enabling the vehicle to
safely drive out of the guideway section. A similar safety feature is used on aircraft.
EVAG also undertook tests to determine the effects of obstacles in the running way;
which proved not to be a problem.
The system has a very good safety record, with only one serious crash in the 25 years of
operation history. The crash was attributable to the guidance system, when the improper
substitution of an experimental aluminum guidance arm for the normal steel guidance
arm led to the guidance arm breaking while the bus was in operation, causing the loss of
steering control and derailing of the bus.
There are no special training courses for drivers, who are informally trained by going out
on the guideway with an instructor. No special certification was required for operation of
the system. A survey of drivers in 1981 indicated that over 95% of drivers experienced
no difficulty using the guideway. The bus was classified as a normal road vehicle
because it has a steering wheel and is able to operate on normal streets, and the guideway
was classified as a private road rather than a public road because other vehicles are not
permitted to use it. Exceptions had to be granted for the extra vehicle width produced by
the protrusion of the guidewheels on the side and for the extra length and weight of the
original dual- propulsion buses.
25
5.3.4. Maintenance
The trailing axle guidewheels typically last about one year, with exceptions if the wheel
breaks off due to hard contact with the curb. The only special maintenance actions
needed on the vehicles are visual inspections of the guidance arms to check for cracks,
and checking the rubber rings on the guidewheels for wear.
The gauge of the guidewheels is 2.605 meters, which is slightly wider than the 2.6 meter
channel. This provides a snug fit for the vehicle and ensures that guide wheel contact on
both sides of the vehicle is maintained. Maintenance of the guideway has been minimal,
however expansion and contraction at the joints between the sections has resulted in
cracking of the concrete. These fractures in the concrete have not caused any long- term
problems and have been easily repaired.
5.4 Observations
The guidance technology for the Essen application seemed appropriate at the time it was
implemented. The guided buses took advantage of old streetcar right- of- way, and were
able to fit within the space available. The locations selected for the guideway application
are in areas where the curbs do not disrupt left turning traffic or other vehicle maneuvers.
However the need to accommodate access and auto turns limited the application to those
discrete sections of the bus route. The need for a vertical surface is the major drawback
of the system; however this same surface also provides a form of safety barrier should the
vehicle try to veer out of control. The mechanical linkages are easy to repair and no
additional staff are required. Although the vehicle cost and complexity are modest, the
tight tolerances on construction of the guidance surface make the infrastructure costs
significantly higher than the electronic guidance systems, comparable to the cost of
streetcar tracks.
The O- Bahn busway in Essen provides accurate and reliable guidance for its buses, but it
does not have the flexibility or smoothness of ride of some of the more modern electronic
guidance systems. The “ bumping” over the joints was clearly discernable, although this
may also be partly attributed to the age of the vehicles used for the system.
The mechanical guidance technology developed 25 years ago has demonstrated
effectiveness for its designed purpose. The most remarkable experience about the
guideway system is the lack of wear and tear on the running surfaces. These surfaces
have been exposed to constant bus traffic for 25 years and there were no visible signs of
deterioration. Nevertheless, the city of Essen decided not to expand its applications.
Worldwide, the technology has only been adopted in a few other locations in Australia
and the U. K. in the past 20 years.
26
6. Magnetic Guided BRT and People Movers
FROG Navigation Systems developed the magnetic guidance system used on the Phileas
bus and the Park Shuttle automated ( driverless) people movers. The delegation visited
GROG in Utrecht, Netherlands.
6.1 Background
The company was started 20 years ago, and its primary products and experience have
been automated guided vehicles ( AGVs) for industrial applications including factory
automation and automated freight transport systems. The company moved into people
moving applications in recent years, which they noted are much more challenging
because of issues of ride comfort, passenger safety and customer acceptance.
The company name is an acronym for “ Free Ranging On Grid”, which is their approach
for defining vehicle paths based on a grid of fixed reference markers. Their original
reference markers were Texas Instruments radio frequency transponders, but since 1998
they have been developing systems based on permanent magnets, which are cheaper, last
longer, permit higher operating speeds, require much smaller sensors on the vehicles,
and use less computer power. Typical FROG guidance systems are designed based on
( 1) mapping layout ( routes, obstacles and stopping points are programmed in the on-board
computer of the vehicles), ( 2) odometry ( measuring the steering angle and counting
wheel revolutions to calculate the positions on the map) and ( 3) calibration ( comparing
the calculated position to the actual position by means of the reference points, e. g.
magnets embedded in the road surface).
FROG designs vehicles and their control systems and integrates systems on the vehicles,
but they do not do the mechanical work of vehicle assembly. They have about 70
employees. They have a modular system design with a “ FROG box” that is used for
lower- level control of vehicle functions, regardless of the type of vehicle ( from factory or
warehouse AGV to container carrier in port, automated people mover or bus), but with
parameter adjustments to accommodate the vehicle differences. They also have a
supervisory control system for monitoring vehicle status and traffic conditions, which
they call “ Super- FROG”. Their overall design philosophy is based on putting relatively
more intelligence on the vehicles and less in the infrastructure, because they believe this
makes it easier to control, maintain and upgrade the system, so that even changes in
vehicle routing can be accomplished as software updates.
6.2 Applications
The initial people mover application by FROG was the Park Shuttle driverless vehicle
guided using TI radio frequency transponders, which they began developing in 1995 to
provide shuttle service between a subway station and a business park called Rivium in a
suburb of Rotterdam ( Capelle an der Ijssel). A similar vehicle was demonstrated at
Demo ’ 98 in Rijnwoude, NL, and then put into public service to provide “ horizontal
elevator” shuttle service in one of the parking lots at Amsterdam’s Schiphol Airport.
These vehicles were prototypes with room for only 10 passengers, and operated for
27
almost 7 years, but have now been withdrawn from service. Those operations provided
useful experience on issues of passenger comfort, safety and reliability.
FROG has also implemented automation for 4- passenger Yamaha golf carts that were
used to carry passengers at the Floriade flower exhibition outside Amsterdam in 2003.
That experience showed that golf carts designed to drive 20 km per day on average were
not robust enough to provide continuous transit service. The guidance system for the
Floriade was an inductive cable follower because it would have been too expensive to
install the grid of magnets that FROG would normally use. The inductive cable follower
system was vulnerable to shut- downs because of static electric charges associated with
nearby thunderstorms.
In addition to the Phileas bus system described in an earlier section of this report, FROG
has also developed a second- generation Park Shuttle driverless people mover, six of
which will be put into service at the Rivium this summer. These vehicles have 12 seats
and can carry up to 15 passengers, in a more robust design than the first generation.
6.3 Experience
The FROG experience is subdivided into three categories, for generic experience with
vehicle guidance, followed by more specialized experience with guided buses ( Phileas)
and with automated people movers ( Park Shuttle).
6.3.1 General guidance experience
FROG’s experience is that selection of sensing technology is the most critical step in
developing guidance technology. They experimented with various sensing technologies.
In the mid 90’ s, FROG developed the people mover system based on Texas Instrument’s
radio transponders. However this approach was abandoned because of the size of the
antenna needed on the vehicle, the cost and limited lifetime of the transponders, and their
limitation to low- speed vehicle operations ( because of computational complexity).
FROG later worked with their French partners to implement a wire guidance system on
Yamaha automated people movers and found that the wire guidance system is
significantly impacted by lightning. They disfavored GPS because of concerns about
reliability and availability, but noted that it could be used to supplement another
technology ( not, however, as the primary guidance reference). They did not think that a
vision based guidance system was sufficiently reliable. FROG later switched to magnetic
guidance for their people mover products. They prefer the combination of magnets and
mapping, which is their approach, but they noted the need for generally- accepted
standards in order for this to become widely adopted. When asked about safety issues
involving electronic vehicle guidance, the response from FROG was an in- depth
explanation of the advantages of magnetic guidance over other technologies.
FROG’s reference magnets for vehicle guidance are installed within an accuracy of 5 cm,
but are then surveyed after installation to identify their actual locations to within 2 to 3
mm. The survey for a one- kilometer demonstration track in Antibes last year took one
28
day. The vehicle is driven manually along the desired path several times, and its position
relative to the magnet locations is recorded. These recorded trajectories are then
smoothed to reduce jerk in order to produce the final desired reference trajectory that the
vehicle is commanded to follow. This trajectory smoothing function appears to be one of
the current problems in the Phileas installation in Eindhoven. The actual vehicle positions
are estimated primarily by odometry ( counting wheel rotations), and calibration
corrections are applied based on the measurements of position relative to the reference
magnets. This is used to continuously update the calibrations to eliminate errors
associated with changes in tire radius ( from tire inflation changes, for example).
Figure 8. Automated Park Shuttle People Mover
Guidance system performance requirements have been defined jointly with customers,
but the driver interface and legal requirements on the systems have been defined by the
customers. For example, the widths of the guidance path as compared to the vehicle are
defined differently for Park Shuttles and Phileas. In the former case, the Park Shuttles
need “ an obstacle free space”, but its width was not explicitly defined. In the case of
Phileas, the bus developer, APTS, asked for a path that would be 20% narrower than a
conventional bus lane.
FROG indicated that their system has already matured considerably through thousands of
vehicle miles of operating experience, but there are additional refinements that they
would like to be able to make to reduce the current four control units per vehicle to two.
29
FROG indicated that safety is the most important part of the design of the automated
guidance system.
FROG’s philosophy is to train customers to be self- supporting for system maintenance
and modification. Maintenance is typically done by the vehicle customers, but their staff
are trained by FROG. FROG provides parts and software updates.
6.3.2 Guided bus systems
FROG believes that rear wheel steering is needed for articulated buses in order to ensure
the tail of the bus stops in parallel to the bus stop curb. FROG indicated that the current
tracking accuracy for the Phileas guidance system is 2- 3 cm. The crab steering of the
Phileas bus was estimated to save 20% of the docking time, as well as reducing the
required length of the docking platform. The advantage could be greater for a system
with off- line stations, by reducing the length and complexity of the access ramps that
would be needed.
When drivers were first exposed to the guidance system, they feared it, from both
technology and job security perspectives. The system gained drivers’ acceptance over
time, as the drivers learned that they could still take control of the bus when necessary.
The transition to automated operation of Phileas is triggered based on detection of the bus
position upstream of a station, after it passes over a coded section of magnets. The driver
can take over control by applying torque to the steering wheel and can re- engage the
automation by pushing a button. However, when under manual driving, driver behavior
changes. Drivers drive Phileas like a car and make sharper turns, which could induce
safety implications when drivers switch back to regular buses.
When asked about feasibility of retrofitting their guidance system onto a normal bus
( rather than a special bus like Phileas), FROG thought it was probably feasible, but at a
penalty in system reliability and robustness. They expressed concern about how to
guarantee the alignment of the rear section of an articulated bus unless it has active
steering of the rear axle. They also noted the need for special sensors to detect the
articulation angle in real time.
Although FROG was not involved in the development of the buses used in the
Zuidtangent BRT service around the southern side of Amsterdam, they noted that it
achieved close station docking tolerances with manual steering of the buses. In order to
do that, the tires rubbed up against a special- shaped curb, however, that has produced
accelerated tire wear and is therefore now being discouraged.
When asked about possible applications to bus maintenance yards, FROG suggested that
an automated “ tug” could be used to pull buses through the maintenance facilities,
analogous to some of their industrial automation applications. In this way, not all the
vehicles would need to be equipped, but only a limited number of “ tugs”.
30
6.3.3 Automated people mover systems
The first- generation Park Shuttle people movers were designed as low- cost prototypes
and were therefore not very refined in terms of passenger amenities, propulsion system or
durability. However, a survey conducted at the Rivium business park showed that most
passengers thought the automated people mover system was friendly and pleasant. For
the Schiphol airport application, because the people who use the system may not be
repeat riders, passengers were generally not familiar with system operation and therefore
some passengers became frustrated with it. FROG has learned from the experience with
those vehicles in developing the second generation vehicles, which were described as
“ more balanced” designs.
Obstacle detection is a major issue for the driverless people mover systems. Initial
obstacle detection sensors were designed to detect all possible obstacles. However, the
obstacle detection system was often tricked into stopping the vehicles by swirling leaves
or rabbits in the path of the vehicles. FROG designed a more intelligent detection for the
second generation people movers, using two Sick laser scanners, which can detect objects
within a range of 40 to 50 m ahead of the vehicle. FROG would like to combine these
with infrared detection of body heat in future systems in order to be able to identify
people with high reliability, to ensure that the vehicles do not hit pedestrians. They
would also like to add side obstacle detection so that the paths of pedestrians approaching
the vehicle trajectory on potential collision courses can be predicted. Additionally, the
right of way for a driverless vehicle needs to be clearly demarcated so that pedestrians
and other vehicles do not unintentionally stray into it. This could be done with curbs,
fences or shrubbery. In the Schiphol application, the places where other vehicles crossed
the path of the driverless vehicle were controlled with crossing gates, while pedestrian
crossings had audible warnings ( particularly since the electrically propelled vehicles
make almost no noise) but no gates.
Video cameras are installed in front of the driverless vehicles, and their outputs are sent
to a control room where the operator can use their information to diagnose any problems
that arise. The cameras face ahead of the vehicle, inside the vehicle and under the vehicle
in order to help the operator diagnose problem causes that could be located in any of
those places. They are also logged on the vehicle to record any incidents that may have
liability implications.
A driverless people mover system still needs a system operator in a control room to
monitor performance and handle unexpected circumstances ( determining if a halted
vehicle should be returned to automated operation) and somebody to handle maintenance,
regardless of whether it involves a single vehicle or multiple vehicles. This indicates a
potential economy of scale for systems with larger numbers of vehicles.
The driverless vehicles have significant user interface challenges, so that users can
readily understand how to use them to get where they want to go, but also to make sure
that users do not mis- use them ( hitting external emergency stop button to catch a
departing vehicle, or deliberately going in front of it to force it to stop). This is
31
particularly important for vehicles that are frequently used by one- time passengers ( such
as at airports), as compared to vehicles used by daily commuters.
For the new generation of Park Shuttles, FROG is trying to define the most appropriate
relationship with the vehicle operator, Connexxion, for system operations and
maintenance. This involves questions of handling immediate repair needs as well as
more general monitoring of system operations, and appears to have significant cost
implications.
6.4 Observations
Although the group did not get a ride in a demonstration vehicle at FROG, we had a
chance to see both first- generation and second- generation Park Shuttle vehicles, to
observe the differences in quality of construction. We also heard about the many
technical lessons learned through the process of developing and refining these vehicles.
It is not clear how applicable such driverless vehicles would be in the U. S., with different
concerns about personal security and liability than in Europe. The guidance system for
Phileas appears to be fully integrated with the overall vehicle design, which might offer
both pros and cons. The integration offers opportunity for synergies to be designed for
different system functions. On the other hand, if not designed properly, when one
component breaks down, several functions can be affected.
32
7. Transit Operator’s View on Guided Buses
Connexxion is the only transit operator in Europe that operates automated guided transit
vehicles as part of their regular services and has accumulated valuable experience. The
delegation visited Connexxion and ANT Consultants in Hilversum, Netherlands. The
purposes for this visit were to learn about ( 1) how the decision is made for deployment of
new guidance technologies, ( 2) their experience with the Park Shuttle people mover
vehicles, and ( 3) how transit systems are operated in the Netherlands and about other bus
operations such as the Zuidtangent.
7.1 Background
Connexxion, which was formed by a merger of several predecessor organizations in
2000, is the largest contract ( concession) operator of public transit services in the
Netherlands ( with 65% to 70% of the market). It is currently wholly owned by the
national government, but they are planning on privatizing it within the next couple of
years.
Only Amsterdam, Rotterdam and Utrecht operate their own public transit systems, while
all others in the Netherlands are contracted out. Connexxion has annual revenues of
about 􀀁 1 B, with 15 K employees and over 1 M daily passengers. They operate 2600
buses, two trains and 27 trams ( only in Utrecht), 2 long distance train lines, 3,125 taxis,
22 ambulances, 310 tour cars, and 16 vessels. The public transit operations account for
about half of the company employees and revenues. Their operations include the Phileas
buses in Eindhoven and the Park Shuttle people movers at Rivium, as well as the
Zuidtangent, which is the largest BRT system in Europe ( 24 km of dedicated bus lane on
a 41 km route, 33 articulated buses, 25 K daily passengers).
7.2 Applications
It was particularly important to understand the relationship between the public transit
agencies and the operating companies such as Connexxion. The public agencies issue
“ tenders” ( essentially RFQs) for contractors to provide transit services, with specified
levels of service required ( in terms of number of passengers carried, schedule adherence
and percentage of scheduled trips operated). However, the process is not yet mature and
they do not have good measures for quality of service.
The public agencies pay for the vehicles and associated infrastructure, so they are the
ones to decide whether to introduce a new system, technology or service. Through
tenders, government may ask for technologies to be included. Because farebox revenues
typically cover only 40% of operating costs, the public agencies also need to provide the
subsidies to make up the difference.
33
The main context for this visit was Connexxion’s role in operating the Park Shuttle
driverless vehicles at the Rivium business park. The implementation of that system was a
joint decision by the local government and the transit operating company at the time ( a
predecessor to Connexxion), but surprisingly no mention was made of the role of the
business park developer or owner.
The Park Shuttle connects the Rivium business park and a subway station in Rotterdam.
Connexxion had been searching for a high quality and economically viable means to
provide connections between the business park and the public transportation station in
order to make public transportation more attractive. In 1995- 1998 the pilot project was
developed, involving 1.5 km of elevated and dedicated roadway and three automated
Park Shuttle vehicles.
The operating speed of the first generation Park Shuttle is 30- 40 km/ hr. It makes 5 stops
with 1.5 minutes waiting time at each stop. Every 15 minutes a shuttle makes a complete
circuit. The capacity is 12 people per vehicle, with hourly system capacity of 500 people.
7.3 Experience
When the Rivium Park Shuttle system was first created, the major investment was in the
track infrastructure, and it was important to make sure that this investment would not be
lost if the automated vehicles did not succeed, so the backup plan was to use it as a
bicycle right of way. The national government provided research funds to support its
financing.
The usage of the Rivium system grew from 200 to 500 people per day during the time it
was in service, and it was generally well accepted by the travelers. However, the scale of
the system was judged too small to be economical, considering the number of passengers
and distance covered, versus the need for full- time operating and maintenance staff.
Also, with only three vehicles, the headways were much longer than the 2 minute
headway between trains on the subway line to which it was providing the connection.
Passengers were initially skeptical about the driverless shuttle, but it gradually became
accepted. In the first phase of operations, availability was 98% ( which was not worse
than normal buses) and there have been no major safety incidents ( a couple of incidents
occurred during the pilot period, but there were no injuries), but that is not considered
good enough for this type of system. People were getting impatient when failures occur
because a failed vehicle can block the track for the entire system and there was no
dynamic information to inform passengers about the failures and the expected arrival
time. Additionally, the speed was not optimal, constrained by a passing point of the
single lane infrastructure. Connexxion decided to shut down the system for an expansion
and to develop the next generation shuttles. The new system is scheduled to start service
in September 2005.
34
The expansion will increase passenger capacity ( up to 2000 passengers/ hour) and the
length of the track ( to 2 km), and will add four secured grade crossings for other vehicles
and pedestrians to the dedicated lane ( with traffic signals and barriers). The new system
will operate on a scheduled basis during the peak periods and on demand in the off- peaks
in order to maximize efficiency of vehicle usage. The peak- period headways between
vehicles will be short enough to reduce the wait time to 1.5 minutes. Since the vehicles
are entirely battery powered, the batteries need to be recharged during the mid- day off-peak
period, when fewer vehicles are needed to provide public service.
The new system will include automated functions at the maintenance area so that the role
of the human operator can be minimized ( door opening, docking for battery recharging,
parking spare vehicles). The improvements were based on experience with the first
generation system and include enhanced reliability, dynamic passenger information about
vehicle arrivals, improved travel speed, reduced wait time and improved trip time
reliability; increased capacity; and reduced maintenance costs.
The capital cost of the expansion includes 􀀁 4.8 M for construction of the two- lane track,
five stations, a bridge extension and garage/ maintenance facility plus 􀀁 2.2 M for six
vehicles, with their hardware and software, but without including FROG’s costs for
development. The financing of the new system involves the Rotterdam regional
government paying for infrastructure, while Connexxion pays for the vehicles and
supervisory system. A couple of government programs help pay for development of the
vehicle system technology.
Annual operating costs for the old system were 􀀁 190 K per year, and for the new system
are expected to be 􀀁 250 K per year plus a 􀀁 300 K depreciation expense. The system is
expected to carry 500 K passengers per year when re- activated, but the regional
transportation authority will have to cover the expected shortfall relative to farebox
revenues. The expected farebox recovery ratio is 67% of operating costs.
The longer- term plan involves no full- time operator on- site, but rather a remote
operations supervision center that would be shared with other transit services for
economies of scale. There still needs to be an on- site support person to handle urgent
service needs, but that will be provided on a third- party support contract on an as- needed
basis.
The reliability of the vehicles varied widely in the first phase. Sometimes, there could be
three or four failures in one day, while at other times one or two months could pass
without failures. This appeared to be affected by the operator’s setup process when
initiating daily vehicle operations. No dominant failure mode or source was identified.
For the new system, if maintenance costs on the new vehicles are higher than specified in
the contract with FROG, they will have to pay a penalty.
The Schiphol Airport project will not be revived because that was determined to not be
an attractive application for a driverless guided vehicle system.
35
7.4 Observations
Our hosts offered several relevant observations at the end of the discussion about their
experiences with the Park Shuttle vehicles:
The decision for moving forward with driverless Park Shuttles was difficult and the
initial decision underestimated the difficulties and the cost. However, the first- generation
FOT proved general traveler acceptance of driverless vehicles.
If the second- generation vehicles at the Rivium are well accepted and are judged to be
successful, this could stimulate interest in using such vehicles elsewhere. However, local
officials are unlikely to decide to try them elsewhere until success is shown here.
The Park Shuttle vehicles are twice the cost of a 12 m standard bus, although they have a
much smaller passenger capacity. When total life- cycle costs are compared, the savings
in driver labor could be an important factor in their favor for applications that require
operations for many hours per day. If the vehicles are produced in large quantity, the
costs could come down significantly.
36
Appendix A
Final Detailed Itinerary
Technical Visit to Europe
Vehicle Assist and Automation Technologies
1/ 24 Monday Meeting with French DOT at Paris- La Defense
Morning Research and policy concerning BRT activities within the
French national research program on transportation
- Policies
- Technologies
- Supply and demand
Speakers: François Perdrizet, director, Research
Administration, French DOT
Jean Orselli, principal engineer , French DOT
Michel Muffat, Research engineer, Research
administration, French DOT
Mathieu Goetzke, Research Administration, International
liaison coordinator, French DOT
Sylvie Niessen, Engineer, Transit Administration, French
DOT
Jean- Luc Ygnace, Researcher, INRETS
Optical guidance systems for buses
Speaker: Paul Edouard Basse, Principal Engineer, Guided
Buses Department, Siemens
1/ 25 Tuesday
Morning Meetings with Rouen Transit
Jean Rince, public transit manager, Rouen transit
authorities
Sebastien Holstein, Technical deputy director, Connex-
Rouen
Afternoon Ride AGORA guided bus
Notes: AGORA is instrumented with Siemens optical guidance
( the same used on CIVIS bus) that performs precision docking
and lane control capability.
.
37
1/ 26 Wednesday Meeting with Phileas Operator
Morning Host: Mr. Theo Dijk, Phileas Project Manager for SRE
( Samenwerkingsverband Regio Eindhoven), the regional
public authority responsible for Phileas bus operations.
Phileas bus ride on initial line, out to Airport, and return to
SRE. DVD presentation on Phileas bus at SRE, and Mr.
Dijk will answer our questions.
Afternoon Meeting with TNO Automotive on Safety Certification 1
Host: Jan P. van Dijke
Senior Project Manager
TNO Science and Industry
1/ 27 Thursday Meet with Essener Verkehrs AG ( Transit agency for Essen)
Morning Meet with Essener Verkehrs AG ( Transit agency for Essen)
( O- Bahn Guided Bus Operators) 2 at the headquarters of
their sister company, ABELLIO GmbH.
Host: Prof. Hans Ahlbrecht
Project Manager EVAG
Information and discussions:
- the Essen transportation system ( Light rail, streetcars,
guided buses, standard and articulated buses)
- the technology of the Essen guided buses
- questions and answers
Afternoon Visit of the guided bus system in operation, riding on the
buses
1/ 28 Friday Visit FROG Navigation
Morning Visit FROG Navigation in Utrecht for briefing and to see
new- generation ParkShuttle vehicle for operation next
summer at Rivium.
Hosts: Carel van Helsingden and Robbert Lohmann
1 TNO Automotive in the Netherlands conducts safety certification of the guided systems in
Europe and has performed safety certification for Phileas bus.
2 O- Bahn mechanically guided bus was the first of its kind and has been in operation for decades.
38
Afternoon Meeting with Connexxion at their office at Marathon 6 in
Hilversum, together with ANT Consultants ( Nicolaas de
Ronde Bresser), who were responsible for hiring
Connexxion. They have had responsibility for operating
the ParkShuttle people mover systems, and will discuss
experience with operations and maintenance issues on
ParkShuttles.
39
Appendix B
Europe Technical Visit Participants
Federal Transit Walter Kulyk
Administration Director, Office of Mobility Innovation
US DOT Federal Transit Administration
ITS Joint Program Yehuda Gross
Office, US DOT Manager, Mobility Services for All Americans Initiative
ITS Transit Program Manager
U. S. Department of Transportation
AC Transit Jim Cunradi, AICP
Manager, Bus Rapid Transit Project
Alameda- Contra Costa Transit District
David M. Angelillo
Planning Operations Administrator
Alameda- Contra Costa Transit District
Lane Transit Graham Carey, P. E., AICP
District BRT Project Engineer,
Lane Transit District,
Stefano Viggiano
Director of Development Services
Lane Transit District
LYNX Doug Jamison
Project Manager, Strategic Planning
Central Florida Regional Transportation Authority
Mitretek Matthew Hardy
Lead Transportation Engineer
Mitretek Systems
SANDAG Dave Schumacher
Principal Transportation Planner, Land Use and Transportation
Planning Department
San Diego Association of Governments
Maurilio ( Mario) Oropeza
40
Senior Project Manager
Mobility Management and Project Implementation Department
San Diego Association of Governments
Brian Sheehan
Transportation and Land Use Planner
San Diego Association of Governments
UC Berkeley Wei- Bin Zhang
Leader, Transit Research Program, California PATH Program
Co- Director, National BRT Institute
University of California at Berkeley
Steven Shladover
Leader, Transportation Systems Group
California PATH Program
University of California at Berkeley
41
Appendix C
Questions Prepared Prior to the Visit
Government Agencies and Transit Operators
Institutional/ Statutory Overview
􀀁 Provide an overview of the acquisition process ( selection and purchasing) for
transit equipment and technology.
􀀁 Does the government provide any incentives to use transit?
􀀁 Who maintains the roadway infrastructure on which the transit vehicles operate
and/ or the technology will be installed?
􀀁 What organization had the authority to modify the infrastructure?
􀀁 What is the relationship and/ or arrangements between bus manufacturers and
technology suppliers on liability indemnification?
􀀁 Are the transit operators and/ or government organizations self- insured or
externally contracted?
o How does the transit agency manage its insurance plan?
Selection and Implementation of Vehicle Guidance Technologies
􀀁 What were the reasons for installing vehicle guidance for transit operations?
o Was a business case made for the technology?
o Was the technology proven to be cost effective before making the
decision?
􀀁 What are the most important concerns that you had when vehicle guidance
systems were first evaluated? Do you still have concerns in these areas?
􀀁 What are the most important factors that influenced your decision to implement a
vehicle guidance system and/ or precision docking system? Such as:
o Reduced infrastructure cost through the reduced lane width requirements.
o Improved system reliability.
o Reduced travel time.
o Increased ridership.
o Reduced dwell time at station/ stop.
o Improving accessibility for persons with disabilities.
o Improving the public perception and/ or image of public transportation
o What problem( s) was it intended to solve.
􀀁 What were the institutional challenges you had to overcome in order to select the
vehicle guidance technologies? Such as:
o Legal issues.
o Liability.
o Labor issues
42
􀀁 What were the technical issues you considered in selecting the specific vehicle
guidance technology you use?
o For station docking, how tight are tolerances for street and sidewalk design and
how has real world experience effected its operation ( drainage, weather, transit
vehicle design, other vehicles that may operate in or near the station area, etc)?
􀀁 What process did you go through in placing the vehicles into service in order to
insure a safe and reliable service? Such as:
o Obtaining approval for using innovative ( non- proven) technologies
o Safety certification.
o Rigorous system testing.
o Public outreach.
o Employee Outreach.
• How did you provide requirements to the manufacturer for the technology? Such
as:
o Agency- developed functional specifications or performance specifications.
o Relied on manufacturer’s knowledge and requirements.
􀀁 What decision, if any, has your agency taken with respect to:
o Whether future procurement of new buses would have automation as part
of the procurement?
o Whether future lines ( or service) would be using buses with
automation? and
o Whether any of unequipped buses now in use will be retro- fitted with
automation equipment?
Operation
􀀁 Explain the process for introducing the technology to the drivers
􀀁 What was the initial impression drivers had regarding the technology when it was
first introduced?
􀀁 How do drivers respond to it now?
􀀁 What kind of driver training is provided in order to for the vehicle guidance
technology to be used?
􀀁 Regarding the overall public image of the vehicle guidance system:
o How do passengers like the vehicle guidance technologies?
o Is there any evidence that ridership has increased because of the vehicle
guidance technologies?
43
o Have there been any type of surveys of passengers regarding the
technology?
o How has the system been treated in the public media ( newspapers,
magazines, radio, TV)?
􀀁 How well does the vehicle guidance technology meet your expectations in terms
of:
o System performance.
o System reliability— mean time between failure; mean time between
working failure
o Maintainability
􀀁 What " surprises" have you encountered because of the technology?
􀀁 How often does the system break down?
o Minor breakdowns that can be handled without removing the bus from
service
o Major breakdowns that require removing the bus from service
􀀁 With regards to system failures and malfunctions:
o What policies and procedures ( for drivers and operations center) have
been established if there is a system failure?
o What course of action should be followed in the event of a guidance
system failure ( fail- safe/ fail- soft)?
o What actions are taken by the driver in the event of a system failure?
o What actions are taken by the control/ operations center in the event of a
system failure?
􀀁 What are the capital, operating and maintenance costs of technology to date?
􀀁 Are there any aspects of the guidance technology ( e. g., tolerance) that you have
deployed but do not need or did not deploy but wish to have?
Benefits
􀀁 In the area of cost- benefit analysis:
o In what respect do you feel the money used for vehicle guidance has been
cost effective?
o How do you evaluate the cost- benefit of vehicle guidance technology?
o How did you measure the benefits of the technology?
o What were the benefits after 1 year in operation?
o How many years do you estimate it will take to break- even between the
cost of the technology and benefits derived?
o What is your estimation of the net benefits over the life of the vehicles?
44
o Do you see electronic guidance as a key element to BRT’s success, or is
this a desirable but non- essential amenity?
􀀁 What were the benefits that you believed the vehicle guidance system would offer
when you selected it?
o Did the technology meet your expectations?
􀀁 How has the vehicle guidance and/ or precision docking technology impacted:
o Travel Time: Running way, station dwell, wait and transfer.
o Reliability: Travel Time, Service.
o System Capacity.
o Connection protection at transfer points.
􀀁 What is your overall opinion of the vehicle guidance technology?
􀀁 Do you believe your agency made the correct decision in choosing vehicle
guidance? Why?
Liability
􀀁 What is the liability implication of the implementation of vehicle guidance
systems?
o How does your agency handle it?
􀀁 What issues did your agency have to address with regard to liability and vehicle
guidance before implementing the system?
o Do you believe there is a difference in liability implication and/ or severity
if an accident is caused by driver or by the vehicle guidance system?
o Do you have any special data recording systems to help determine the
causes of accidents?
o Will the cost of a guidance failure be passed back to the manufacturer?
o Were any special agreements or legal changes needed?
􀀁 What is the process involved in determining countermeasures to reduce liability
exposure?
–
Safety certification:
􀀁 How do you define safety?
􀀁 What are your safety action plans?
o What happens when a vehicle guidance system fails?
o What happens when the vehicle guidance system shuts down or is
inoperable?
45
􀀁 What level of safety certification did you go through?
o Can you provide an overview of the certification process?
o What were the lessons learned from this process?
􀀁 Who is involved in safety assessment and certification before equipment is
deployed?
Maintenance
􀀁 Please provide an overview of your maintenance policy and/ or program.
o What repairs do you take care of in- house and which do you outsource?
o What types of training programs are carried out for maintenance crew?
o How often are maintenance crews trained?
o Is training conducted through external institutions or internal programs?
􀀁 What is the size of your vehicle fleet?
o Conventional buses
o Buses with guidance systems
􀀁 What is the size of your vehicle and equipment maintenance staff?
o For entire bus fleet
o Specifically added to handle guidance systems
􀀁 What changes to the maintenance system were required because of the vehicle
guidance technology?
o Significant changes in maintenance procedures
o Purchase of additional maintenance tools and equipment.
o Expertise, training, knowledge.
o Expense with regard to preventive maintenance
o Diagnosing system problems
o Infrastructure ( e. g., lane markings)
o System calibration.
􀀁 What is the current frequency of:
o Preventive maintenance and repair work on major bus systems prior to
vehicle guidance technology.
o Preventive maintenance and repair work on major bus systems after
vehicle guidance technology.
􀀁 Is the maintenance of vehicle guidance systems similar to or very different from
the maintenance of other electronic vehicle technologies?
46
Deployment and Institutional Issues
􀀁 How has your agency worked with other transportation organizations ( transit
agencies, planning organization, local DOTs) in your local and/ or regional area on
projects relating to the implementation of new technologies or new services?
o How important to the success of these projects is inter- jurisdictional
coordination and communication?
􀀁 How easy or difficult is it to sell or market the changes associated with bus rapid
transit ( BRT) and vehicle guidance systems?
o Have special or customized approaches been taken such as educating
motorists and pedestrians on interacting with other vehicle guidance
system operators?
􀀁 How do the passengers that ride your buses or local/ regional media currently view
your agency in terms of the service it delivers, its current performance and its
reputation?
o Has the vehicle guidance technology changed the current image?
􀀁 What is the level of passenger acceptance of buses operating with vehicle
guidance systems? Was it immediate or was there a need for a “ breaking- in”
period that could offer lessons learned for implementation of new systems, such
as transit lane assist?
􀀁 Has there been any discussion or deployment of vehicle guidance technologies to
improve the daily maintenance operations of buses? ( washing, refueling,
cleaning)
47
For the Meeting with TNO Automotive
􀀁 What is your process for safety certification?
o How do you approach the safety certification process?
o How are requirements developed?
o How are these requirements measured and met?
o How are new systems, one that has never been tested before, addressed?
􀀁 How much can be done “ on paper” by reviewing design
documentation and software?
􀀁 How much must be based on testing the actual systems?
􀀁 What do you think the main issues are for safety certification, based on your
experience?
􀀁 What kinds of performance and safety problems have you encountered in the
vehicle systems you have worked on?
􀀁 Based on your experience with the systems you have worked on, how would you
do things differently in the future?
􀀁 What do you recommend to system developers in order to minimize the
difficulties they are likely to encounter in attaining certification for new vehicle
guidance systems?
􀀁 To what extent are your certification processes based on specific national or
European regulations, and to what extent are they based on more broadly
applicable engineering and safety principles?
􀀁 How mature is the certification process for automatically guided vehicles? Do
you expect it to change significantly within the next 5 or 10 years? If so, in what
ways?
48
For French DOT
􀀁 Please provide an overview of the French DOT role in local transit operations and
the decision- making process with regards to:
o Purchasing new vehicles/ technologies
o Planning
o Other
􀀁 What is your vision on the development and deployment of new technologies for
transit operations?
o What is the role of the French DOT ( vs local transit agency vs. private
sector) ?
o How is a research and development plan for new technologies created,
funded and implemented?
o What is the role of new technology within transit operations?
􀀁 What are the deployment status and issues related to vehicle guidance
technologies?
􀀁 We understand that Clermont- Ferrand was using CiViS with guidance system but
stopped the service recently. What were the reasons for stopping this service?
49
For the Meeting with Siemens and Frog
􀀁 What are the most common requirements from transit operators when they want
to purchase your vehicle guidance/ automation technology?
o Among these, which are the most critical requirements from
manufacturer’s perspective?
o Have they wanted some capabilities that are not possible to provide?
􀀁 Is the electronic guidance technology tailored to a specific vehicle or can be
applied to any conventional bus?
􀀁 What feedbacks did you receive from the customers?
􀀁 What is your market projection for transit vehicle guidance, precision docking and
automation in Europe and other countries?
􀀁 What improvements to the current technology do you plan to make?
􀀁 Who makes the decision to choose a special guidance system such as yours for
use in a specific transit system?
o Local public agency
o National government
o Transit operating company
o Bus manufacturer
􀀁 What is your relationship with the bus manufacturer?
o What is the boundary in responsibilities between your company and the
bus manufacturer?
o How hard or easy is it to define that boundary when a project is started?
o How do the local transit authorities get the technology installed on their
buses?
􀀁 What is the estimated life of the vehicle guidance technology?