Transcript Driverless Car Summit 2012

Driverless Car Summit
2012
Association of Unmanned Vehicle Systems International
(AUVSI)
Richard Bishop, Automotive Lead
S
AUVSI
S Non-profit industry association focusing on air/land/water
unmanned operations
S 40 years
S 500+ corporate members
S expanding into civilian vehicle autonomy
S Driverless Cars 2022 Initiative: stretch goal
S what does it take to get there?
S DCS 2012 first meeting of its kind
Driverless Car Summit
S Detroit, June 12-13, 2012
S Agenda included:
S White House Office of Science and
S
S
S
S
Technology Policy
USDOT (ITS JPO and NHTSA)
USDOD
Nevada Dept. of Motor Vehicles
Google, GM, Daimler, Continental
S ~ 250 attendees
S www.auvsi.org
Types of Automation
S beyond automated longitudinal control (ACC) to add some level of
automated lateral control (lane centering)
S driver must still maintain vigilance as to any unusual situation on the
road
S How successful will these systems be in maintaining driver
engagement?
S Will driver monitoring become common?
S Traffic Jam Assistant (full control below a speed threshold on
highways)
S automated urban “citycars”
Industry Representation
S Car industry
S Chrysler, Daimler, Ford, GM, Honda, Nissan, Toyota,
Volkswagen, and Volvo Cars
S Suppliers
S Valeo, Continental, others
S Google
S Ground vehicle robotics firms
S Transportation engineering firms
Car Industry Perspectives
S the advent of vehicle automation is by now a given
S systems are being designed to handle the roads and traffic “as-is.”
S The role of the infrastructure is open, as to how traffic
management and mobility enhancement systems adapt.
S Expanding mobility for the disabled and elderly is a motivator
S compelling given the aging of the Baby Boom generation
Car Industry Perspectives
S Dr. Gary Smyth, General Motors
S significant vehicle evolution in the next decades
S expansion of the DNA of today’s personal mobility vehicle to
S
S
S
S
S
include electrification, electronics, and connectivity.
“transferred control” (hands/feet off) by mid-end decade
autonomous driving by the end of the decade.
Super Cruise (combined lat/lon control): near term
Knowledge of the driver state and vehicle capability is essential
investing significant effort into understanding transfer of
control – this has to be done successfully in 2-3 seconds
Car Industry Perspectives
S Dr. Luca Delgrossi, Daimler:
S stepwise approach to automated driving
S looking at autonomous driving as their final goal
S F800 prototype: Traffic Jam Assistance
S requires driver to touch the steering wheel at regular intervals to
stay engaged
S Christian Schumacher, Continental
S combined longitudinal and lateral control
S using equipment currently on the car
S implementing new features through sophisticated software
Industry Perspective
S Chris Urmson, Google:
S key priority is to program the vehicle to
operate as if driven by a human.
S strong focus on defining performance
metrics
S Re introduction: “the perfect is the enemy
of the good”
S we should not wait for perfection
Government
S States passing driving laws
S USDOT assessing risks and addressing challenges
S USDOD beginning deployments for specific vehicles and
operational environments
S Active collaboration between these DOT and DOD being
explored
USDOT
S John Augustine, USDOT ITS Joint Program Office
S “the technology is feasible and deployment achievable.”
S automation can help across the board in addressing road transportation
goals
S JPO funding multi-agency Automated Vehicle Exploratory Research
program
S
passenger cars, heavy trucks, and transit
S John Maddox, NHTSA
S newly defined automated driving research program
S stake in the ground: autonomous vehicles must be “better than humans.”
S goal for automated vehicles: “crash-less … I don't think people are willing
to accept robotic error resulting in killing people.”
S “challenges can be met”
USDOD
S Dr. Jim Overholt, US Army Tank and Automotive Research,
Development, and Engineering Center (TARDEC), Ground Vehicle
Robotics Innovation Center
S “I want to allow soldiers to do something else while doing the
mundane task of driving.”
S Autonomous Robotics for Installation and Base Operations (ARIBO)
program
application of autonomous vehicles to real-world needs at military bases
roads and facilities at these bases can serve as a more “protected”
environment compared to the open road
S Army sees these bases as a good testbed for autonomous operations
S possible way to collaborate with USDOT for testing
S pursuing the use of robotic vehicles at Fort Bragg to ferry “wounded
warriors” to medical care facilities on-base
S
S
User Panel: Urban Complexity
S constant interaction outside the car – other drivers, bicyclists,
pedestrians
S cyclists must be confident they are “seen” in entering an
intersection, and they confirm this by looking at the driver
S how will the self-driving car “communicate” in such a situation
and share the road space appropriately?
S will this new technology be designed for use by the blind to
expand their mobility?
S a series of demonstrations and public campaigns will be important
to inspiring public confidence in automated driving.
Legal Issues Panel
S change the legal infrastructure to enable automated vehicles, or
change the vehicles to adapt to existing law?
S general response: vehicles will adapt to the law – not realistic to
seek for laws to be passed just to address liability issues with
automated vehicles.
S on-board data recording important
S use electronic discovery of evidence to prepare the way today for
lawsuits tomorrow
S using data to prove in court when the vehicle is not responsible for
a mishap.
S significant uncertainty in the legal realm is inhibiting the industry.
Connectivity <> Automation
S connected vehicles and automated vehicles are separate
topics yet inter-related and complementary
S in the longer term, exchange of data between vehicles
promises to enhance performance overall.
S cybersecurity is a significant concern
DCS12 Summary
S interactive format deepened understanding and cross-
industry connections
S telling the story important -- we need a strategy and plan in
educating the public
S cross-industry issues should be addressed, for example:
S testing / certification
S state-level regulations
S continue regular information sharing
2013 Driverless Car Summit
S Detroit
S May or June 2013
S Annual meetings planned in following years