1. Introduction

   The fifth annual meeting of the International Task Force on Vehicle-Highway Automation convened at the Sydney Convention Center in Sydney, Australia on October 4, 2001. The meeting was hosted by ITS Australia, which provided superb administrative support. The meeting was co-chaired by Richard Bishop, Bishop Consulting, and Dr. Steven Shladover, Deputy Director of the California PATH program (Dr. Shladover replaced Mr. Greg Larson, Chief of the Office of Advanced Highway Technology at Caltrans, who was unable to attend).
   The meeting was attended by 30 invited participants, with representation from Australia, the European Commission, France, Germany, Japan, Korea, The Netherlands, the United Kingdom, and the United States. Of this group, 17 are government officials or with AHSRA, 5 are with research laboratories, and 8 with the private sector. One European vehicle manufacturer was directly represented, with Japanese manufacturers represented through AHSRA.
   The meeting was structured in two parts. During the evening of October 4th and the morning of October 5th, overview presentations provided the latest status on projects and policy developments within their programs. The afternoon of October 5th consisted of examination of key issues identified during the discussions which followed the presentations. Going into the meeting, the following issues were seen as primary:
• Deployment Paths
• User and Societal Issues
• Role of the Infrastructure
   This article will provide summary overviews of the presentations given by representatives from Japan, the United States, and Europe. It will then selectively take up the major problems raised in discussion.

3. Summary of Presentations (Facilitated by Dr. Shladover)

(1) The Current Situation and Future Preparation on Development of AHS
    Mr. Shigenobu Kawasaki,
    Director for ITS R&D Division
    National Institute for Land and Infrastructure Management
    Ministry of Land, Infrastructure and Transport
Mr. Kawasaki focused on three areas:
• The flow of research and development on AHS
• Division of roles between vehicles and roads
• Scenario for deployment of AHS

Flow of R&D on AHS
    The initial focus is on improving safety, to gain a 50% reduction in accident fatalities. This is done by focusing on the seven services defined and tested for Joint Tests. The proving tests in 2000 provided the opportunity judge the effectiveness of the services and overall system performance. This has resulted in proposals for practical application systems, which will be followed by field operational tests, leading to Phase One deployment to begin in 2003.

    The field operational tests will focus on:
   • Judging the effectiveness of the system
   • Specification development of infrastructure systems
   • Assessing system reliability
   • Determining the deployment pattern
   • Analyzing cost effectiveness
   • Developing ways to promote the spread of in-vehicle units
    This process will be followed by a Phase Two which defines additional systems and follows the same approach.

Division of Roles between Vehicles and Roads
    The Japanese approach is quite straightforward. For hazards detectable within the capabilities of the vehicle, autonomous vehicle systems are relied upon. For hazards outside the capabilities of the vehicle, cooperative vehicle-highway systems are planned. Infrastructure-based hazard detection is planned.

Considerations for AHS Deployment
    Two scenarios were presented:
   • Cooperative vehicle-highway systems are introduced in some areas, and more areas are gradually added until the whole country is covered,
   • The infrastructure for collection of information is laid throughout Japan, and then cooperative vehicle-highway systems are introduced

    Related to these scenarios, Mr. Kawasaki made two points:
   • Cooperative vehicle-highway systems need to be deployed at the same time for the spread of in-vehicle units
   • The infrastructure for collection of information is one aspect of AHS which can also be applied to road management.

    As an example, he noted that 14 cooperative support systems for the prevention of collision with forward obstacles have already been deployed in Japan, at location where the sight distance is inadequate for drivers to perceive hazards themselves.

    Regarding the provision of in-vehicle information displays, MLIT research has shown that the braking reaction rate for drivers improves 83%, compared to only 57% for roadside warnings.

    The MLIT scenario for deployment of AHS calls for the following steps:
   • Deployment of infrastructure sensing and using this information for road management
   • Providing infrastructure-based services via deployment of roadside information boards (which will raise driver awareness)
   • Deployment of Dedicated Short Range Communications
   • Availability of Cooperative Vehicle-Highway Services

Finally, major challenges were noted as:
   • Verifying the effectiveness of information board services (infrastructure-based)
   • Developing infrastructure for application in both AHS and road management
   • Promoting the spread of in-vehicle units along with the nationwide introduction of DSRC

Discussion
   • Question: Why buy DSRC for in-vehicle warning when it is free on roadside sign? What is the consumer motivation?
   • Question: Do we have a situation in which safety devices are only available in high-end cars, and therefore only for the rich?
   • Question: Can the equipment used to implement the Japanese systems be used to do Intelligent Speed Adaptation (ISA)?
   • Question: Will the Japanese system provide information on recommended speed?
   • Question: Is MLIT focused at all on relieving traffic congestion, or is the focus exclusively on safety?
   • Question: How do you propose to handle the discrepancy between the infrastructure time frame (long) and the vehicle time frame (relatively short)?

(2) ADASE II - Efforts Towards Active Safety in Europe
   Mr. Berthold Ulmer
   DaimlerChrysler AG
   Manager Research and Technology
   Representative Office for European Affairs

    Mr. Ulmer presented information regarding the ongoing Advanced Driver Assistance Systems Europe (ADASE) program. ADASE is funded by the European Commission, with participation from the European Council for Automotive R&D (EUCAR). ADASE II began in August 2001 and will continue for three years.
   ADASE provides overall coordination of EU projects in the intelligent vehicle-highway area. This "cluster" is composed of the following projects:
   ADVISOR
   CARSENSE
   CARTALK
   CHAMELEON
   CHAUFFEUR
   COMUNICAR
   NEXTMAP
   PROTECTOR
   RADARNET
   ROADSENSE

In the ADASE "roadmap," a key output of phase one of the project, driver assist functions span the range from today's systems to automated driving.

"Mainstream" functions are noted as:
   • Stop & Go ACC
   • Rural Drive Assistant
   • Urban Drive Assistant
"Spin-off" functions are noted as:
   • Lane Departure Warning System
   • Road Departure Warning System
   • Front Collision Warning
   • Parking Assistant
   • Blind Spot Monitoring
   • Lane Change / Merging Assistant
   • Crossing Assistant
   • Trajectory Calculation

   The ADASE architecture includes vehicle functions, vehicle-vehicle communication, vehicle-infrastructure communications, infrastructure sensors, and environment sensors (including maps).
    ADASE will operate through Cluster Management to involve vehicle manufacturers, experts, EuCAR, ERTICO, and others.

    Expert workshops will play an important role. Workshops will be held on:
   • Legal aspects and human machine interface
   • Architecture and technology roadmap
   • Road infrastructure design and road-vehicle, vehicle-vehicle communication systems and applications
   • Sensor technologies
   • Effects on safety, throughput, and comfort

    From its position as a central coordinator, ADASE is focusing on key strategies, roadmaps, technologies, public awareness, and preparation of standards.

   Mr. Ulmer made the following additional comments:
   • global marketability of ADAS systems is desired, but at minimum European-wide marketability is wanted.
   • Europe will move more towards infrastructure support for ADAS. For systems with autonomous sensing only, we will come to the point that we spend significant money for only a 1% improvement; instead we should put more intelligence in the road
   • ADASE II is looking at the potential for ADAS systems to improve traffic throughput, and results will feed into the European Sixth Framework program. Projects and field operational testing based on using ADAS for throughput improvements is expected in 6FW.

(3) Ten Year National Program Plan for ITS in the United States & Intersection Safety through Infrastructure and Cooperative Technologies
   Ms. Toni Wilbur
   Technical Director for Operations, Office of Operations R&D, Federal Highway Administration
   US Department of Transportation

    As Ms. Wilbur was unable to travel to Sydney, Mr. Bishop presented her presentation.

ITS Ten Year Program Plan
   The ITS Ten Year Program Plan was requested by Congress in TEA-21 legislation. Its scope includes ITS R&D and definition of programs and projects within both 5- and 10-year timeframes.
   The basic principles of the plan are that it is both a research agenda and a program plan. Recommendations are to be "bold but reasonable." The intent is to show a readiness to undertake significant challenges, but maintaining a solid grasp on practical reality, particularly constraints on time, human resources and money. Most of the challenges relate not to creating new basic technology, but to the need to change and update existing institutions.

   Ms. Wilbur's presentation noted that the application of computers and communications to complex systems like transportation generally moves through two major stages:
   • Stage 1 - applies technology to a specific tasks, but without changing their character or basic sequence.
   • Stage 2 - whole new approaches to solving problems and conducting business begin to appear because the tasks are no longer tied to physical workstations and paper flows.

   For instance, in banking, stage 1 replaced manual ledgers and mechanical accounting machines with computer transactions and magnetic tape. Stage 2 introduced electronic funds transfer, national and international networks of ATM machines, and online, paperless banking.
   The first 10 years of ITS (since 1990) have been Stage 1 processes.

   USDOT goals for 2011 are:
   • to reduce transportation-related fatalities 20%
   • to save $20B per year by improving throughput and capacity
    The Plan is structured into Programmatic Themes and Enabling Themes.

Programmatic Themes:
   • Integrated Network of Transportation Information
   • Advanced Crash Avoidance Technologies
   • Automatic Crash Incident Detection and Response
   • Advanced Transportation Management

Enabling Themes:
   • Creating a Culture of Transportation System Management and Operations
   • Public Sector Roles, Relationships, and Funding
   • Private Sector Roles, Relationships, and Funding
   • Human Factors

   Of interest to ITFVHA, elaboration on the Crash Avoidance theme is as follows:
Outcome: Elimination of large number of crashes. Development, integration, and deployment of a new generation of in-vehicle electronics, cooperative vehicle-infrastructure automation, and selective automated enforcement.
Challenge: System integration, information requirements, human factors (user interface, compensation & acceptance, consistency), antitrust and liability, public acceptance
Research: International program on driver behavior; integration of in-vehicle electronics, acceptable level of surveillance for automated enforcement, attitudes towards driver qualification technology, costs and benefits
Program: Outreach and education, standards, incentives and subsidies for safety technologies
Institutional: Adjustment to antitrust rules to allow pre-competitive research, risk containment on product liability, relationship between public and telematics industry

   Also of interest, the theme Advanced Transportation Management includes a sub-theme of "Advanced Transportation Automation Systems" which includes automated trucks and cooperative vehicle-highway automation:
Outcome: Intelligently and adaptively manage the flow of vehicles (automobiles, public transit vehicles, and trains) across multiple jurisdictions and modes. Traffic management systems will take an active role in controlling selected vehicle types, increasing capacity and operational efficiency
Challenge:Policy and institutional vision, system integration, predictive and evaluative abilities, vehicle-infrastructure interface, public perception and legal liability
Research: Development of surveillance and detection tools, roadway control algorithms, adaptive control, predictive modeling, architecture and standards, costs and benefits, DSRC
Program: Establish funding programs and incentives for partnerships and regional programs, integrate private IVI and public infrastructure, establish linkages to Department of Defense and Department of Energy research programs.
Institutional: Encourage dialog among and between relevant players, address liability issues

As part of examining deployment paths for vehicle-highway cooperative systems, FHWA has estimated market penetration rates as follows:

   The envisioned deployment path for ICA is as follows:
   • Deploy infrastructure countermeasures at high accident locations
   • Add roadside-vehicle communication beacons
   • Vehicle manufacturers offer cooperative features for sale, consumers choose to buy
   • Continue to deploy countermeasures at additional intersections

   Status of infrastructure-based ICA research is as follows:
   • Analyze crashes  =>  Report in June
   • Communication (DSRC)  =>  Project started in February
   • Location & positioning  =>  National Differential GPS (ongoing)
   • Roadway databases  =>  EDMap started in summer 2001
   • ICA systems  =>  IC Projects started in September
   • ICA Human Factors  =>  FHWA Safety R&D projects

   The IVI Infrastructure Consortium was reviewed.
   • Organized as a Pool Fund in 1998 to advance research in IVI Specialty Vehicles
   • FHWA requested transition in 1999 to focus on vehicle highway cooperation to improve safety
   • Investigating intersection collision avoidance and roadway departure crash avoidance
   • Members: California, Minnesota, and Virginia
   • Will complement research to improve traffic operations efficiency, sponsored by Cooperative Vehicle Highway Automation System pool fund

Discussion
   Mr. Bishop and Dr. Shladover elaborated on the situation in the U.S.
   • Development of cooperative vehicle-highway automation systems was a fifth Programmatic Theme in earlier drafts of the Ten Year Plan; it was subsequently removed from this level and the topic was "distributed" through the other topics, such as Crash Avoidance and Advanced Transportation Management. Some believe that CVHAS has not been given enough significance in the current plan.
   • While use of cooperative vehicle-highway automation systems for improvements in traffic flow efficiency is present in the program plan, the structure of USDOT and most state DOT's is not currently conducive to this new direction. Freeway managers don't think about vehicle control as being within their "toolbox," and vehicle-oriented programs are focused strongly on safety only. It is necessary to make new linkages.
   • Question: What is the time horizon for the ten year plan?

(4) California's Vehicle-Highway Automation System Research
   Mr. Greg Larson
   Chief, Office of Advanced Highway Systems
   Division of New Technology and Research
   California Department of Transportation
   Dr. Steven Shladover
   PATH Deputy Director
   University of California, Berkeley

   Dr. Shladover provided this presentation. He began by noting that the transportation system, in effect, is a three-way relationship between people and goods, vehicles, and the roadway infrastructure, with information flowing among all nodes.

   At PATH, Vehicle Development and Testing is underway for
   • Snow removal vehicles
   • Transit buses
   • Heavy trucks
   • Passenger cars

   The PATH Advanced Vehicle Control and Safety Systems (AVCSS) program goal is to facilitate/accelerate deployment of AVCSS in California to help relieve congestion and improve safety, efficiency, and environmental impacts.
   Objectives were presented as:
   • evaluate relative merits of different technical solutions
   • optimize systems to solve California problems
   • integrate vehicle and infrastructure elements to find best mix
   • demonstrate technical feasibility
   • address societal and institutional issues

   PATH AVCSS research directions are:
   • Continue progress towards AHS
   • Concentrate on abnormal/fault conditions and deployment staging
   • Develop and demonstrate truck and bus automation capabilities
   • Develop answers for skeptics
   • Identify and develop useful precursor systems
   • Participate in USDOT's Intelligent Vehicle Initiative
   • Apply AHS experience to other domains

PATH has ongoing projects in:
   • Automation concepts and evaluation
   • Enabling technology development
   • Vehicle control systems development and testing
   • Vehicle safety, fault detection, and fault-tolerant control
   • Energy and environmental implications of automation
   • R&D to support Bus Rapid Transit systems
   • Includes SmartBRT simulation development
   • Rural field test of RoadView snowplow guidance system
   • Development of advanced rotary plow
   • Evaluation of sensor-friendly vehicle and roadway systems
   • Transit bus forward collision warning system requirements definition
   • Transit bus rear collision warning system display
   • Threat assessment simulation development for the General Motors ACAS operational test
   • Data fusion for ACC systems
   • Rear-end collision warning design and evaluation

   PATH is involved in IVI Infrastructure Consortium Intersection Decision Support project. The goal of the project is to improve driver decision-making at intersections and reduce the crash frequency and severity. Infrastructure and vehicle-infrastructure cooperative solutions are under consideration. Activities are focusing on requirements definition, enabling technology development, countermeasure trade-off analyses, subsystem testing, and public demonstrations.

   PATH is also a member of the Cooperative Vehicle-Highway Automation Systems (CVHAS) pooled fund study, which includes eleven state DOTs, USDOT, and Honda. The CVHAS project goals are:
   • To establish a program that addresses traffic congestion and mobility issues via cooperative intelligent vehicles and infrastructure.
   • To identify and conduct research that will advance development and deployment of CVHAS systems
   • To demonstrate the commitment of key stakeholders to advancing CVHAS

   CVHAS project activities are:
   • Case studies of truck and bus automation in Chicago and Tacoma
   • Developing a marketing, education, and outreach program
   • Soliciting additional members -- state DOTs, private sector, and international
   • Promoting inclusion of CVHAS in the 10-Year ITS Program Plan and Research Agenda
   • Submitting language on CVHAS for TEA-21 Reauthorization
   Upcoming Demonstrations
   Extensive work is currently underway for Demo 2003, for buses and trucks.
   The intent is to showcase the benefits of truck and bus automation in the near to middle term. The vehicle demonstrations will consist of platoons of 3 buses and 3 tractor-trailer trucks. The dates are August 16-20, 2003; the demonstration will occur on the I-15 carpool lanes in San Diego (the same ones used in Demo '97).
   A Demo 2006 is also planned, which will be focused on passenger cars. It will address later stages of CVHAS deployment. Multiple platoons of automated light duty vehicles will be operating. Under some conditions, attendees will be able to sit in the driver's position.
   Videos were shown of the bus forward collision warning system, automated truck testing, and automated merging of passenger cars.

Discussion
   • Question: If more dedicated lanes are needed for the PATH AHS vision, then more construction is needed, but the statement was made that it is impossible to do more construction in California. Please explain
   • Question: Are deployment sites for CVHAS and VHA selected within the state? Is Caltrans seriously considering deployment?

(5) Issues Discussion
   The following issues were identified during the course of earlier discussions:
   • Progression to higher levels of automation: smooth, or will there be a break point?
   • Different needs lead to different solutions (like Cybercar); thus there are impacts of urban form on the selection of solutions for different locations
   • Could developing countries leapfrog straight to community-based shared public vehicles, and avoid the mass of private cars?
   • Liability - how will this affect deployment?
   • Deployment trade-offs between mixed traffic operation and dedicated lane operation
   • Institutional structures for using vehicle control to address congestion (many government DOTs are not organized for intelligent vehicle R&D and highway management R&D to come together)
   • Social equity - early safety systems coming on high-priced cars, but some would argue that safety should be available to all
   • Observation: a (more or less) common technology base (vehicle control, communications, navigation) exists for many diverse applications (Japan SmartCruise systems, Intelligent Speed Adaptation, Intersection Collision Avoidance)
   • Architectural issues for AVCSS/ADAS must be addressed, including functional allocations between vehicles and the infrastructure
   • Opportunities offered by Low Speed Automation
   • Degree of safety benefits offered
   • Usage of AVCSS/ADAS systems - are there situations where they should be used, and other situations where they should not?
   • International cooperation needed
   • The potential for the European programs ADASE and ATLANTIC to enhance international cooperation and exchange
   • Need a better definition of User Acceptance
   • Need a comprehensive roadmap for AVCSS/ADAS systems
   • Need integration of system development activities across expertise area, policy issues
   • A workable societal framework is key

   It was noted that there is significant activity in the UK, France, Germany, for instance, but how does it all come together? There is a need to converge. Mr. Ulmer noted that the intent of ADASE II is to bring national programs (in Europe) together in consultation. His vision is to also create links with Japan, the U.S., and others, towards a global ADASE. Mr. Bishop noted that the ATLANTIC project can contribute, as well - ATLANTIC is established to create dialogue on research issues and priorities between Europe and North America and could eventually be extended worldwide. The multi-state CVHAS project in the U.S. is also seeking active participation by organizations on a worldwide basis.

   It was noted that California, the UK, and Germany all have a strong focus on using vehicle control systems to achieve gains in traffic efficiency, which many see as a key future direction for AVCSS/ADAS. These are shown in shading in the table. However, this grouping is based purely on "vision;" levels of investment vary dramatically at this time. The UK is currently considering an active program in this area, whereas California and Germany have multi-year funded programs underway. These are the PATH program in California and the INVENT program in Germany.

(6) Conclusions
   This year's ITFVHA meeting can be summarized as follows:
   Ordinary drivers long for a solution to the everyday headache of traffic congestion. Just as air pollution undermines the public health, traffic congestion degrades the quality of our lives. ITFVHA is the only organization in the world that convenes international conferences to discuss the development of cruise-assist vehicles as a way of addressing the problem of traffic congestion, together with the other advantages, issues, and activities involved in realizing this vision. We are standing at a critical juncture in this process. Working on the assumption that passenger vehicles equipped with cruise-assist will emerge within the coming decade, then we should complete the basic supporting research by the year 2010 so that industry and government can engage in their respective roles and do so in a proactive manner.
   It also became clear that we must establish a program like ADASE on a global level. This will be crucial for our considered study of common global research themes directed toward cooperative vehicle-highway systems and automated vehicle-highway operations of the future.