The Future of Driverless Vehicles (Roundtable)
Jeffrey Miller, IEEE member and associate professor of engineering practice at the University of Southern California, contributed this article to Live Science's Expert Voices: Op-Ed & Insights.
Three years ago, Nissan was the first car manufacturer to announce they would have driverless vehicles ready for consumer adoption by the year 2020. While consumers, and even some experts in the field, noted that this was an aggressive timeline, it didn't seem like an unattainable goal. Void of personal and professional opinions, this announcement did a great service for the driverless vehicle industry, promoting awareness of this emerging technology. Awareness is one of the most important elements in driving this industry forward — consumers aren't going to trust what they don't know, even if the technology has been validated.
In late August, IEEE —the world's largest professional organization of engineers — hosted a roundtable at the University of Southern California to discuss the current condition and future development of the autonomous vehicle industry. The roundtable featured experts from a variety of disciplines, including technology, policy/regulation and law, where we addressed comprehensive industry considerations.
Along with myself, the participants included:
Justin Pritchard — moderator; transportation reporter for the Associated Press
Wei-Bin Zhang — research engineer and a program manager for the California PATH Program and Institute of Transportation Studies of University of California at Berkeley
Bernard Soriano — deputy director, California Department of Motor Vehicles
Bryant Walker Smith — assistant professor of law at the University of South Carolina
A new vision for "seeing" the world
One area that is continuing to grow and will play a large role in the further development of autonomous vehicles is Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communications.
Currently, self-driving vehicles are guided by computer vision technology — whether it's Lidar/Ridar (laser or radar technology) or camera-based sensing — when operating on public roads. However, V2V and V2I are communication methods that will completely transform how vehicles will "see" the road and interact with its environment. Both V2V and V2I are dedicated short range communications (DSRC) devices that work in the 5.9GHz band, have a range of approximately 1000m and can support private data communications as well as public.
At the rate the industry is moving, we'll start to see V2V/V2I become integrated and tested in controlled settings within the next three to five years, but the technology will require constant evaluation before being available to consumers. Although driverless cars will be on the market by 2020, they will not be able to leverage V2V or V2I until a few years later. [5 Ways Self-Driving Cars Will Make You Love Commuting]
V2V and V2I communications will have large-scale benefits that reach beyond the vehicle. Such communication practices will allow for much safer travel by allowing vehicles to be in constant communication with each other as well as their environment, which will greatly reduce accidents and fatalities. Last May, the Associated Press reported on a National Highway Traffic Safety Administration study that found traffic accidents cost the US $871 billion a year — these communication platforms can greatly reduce this number. As a result, traffic patterns and road congestion will also be aided and vehicles will be able to travel at a much faster rate of speed and eventually render traffic signals irrelevant.
Key to implementation will be a high penetration rate of vehicles able to communicate with each other. This will enable self-driving cars to access further data and information regarding their environment, and will work in harmony with already available sensing technology (radar or video cameras). For example, when a vehicle is coming up to a blind intersection, a vehicle in the perpendicular direction could alert other vehicles to whether it will be able to stop as a signal changes.
The next five years will be important in addressing concerns and barriers to the implementation of V2V and V2I. In August, the National Highway Traffic Safety Administration (NHTSA) issued a release that announced proposed rule-making and an initial analysis of V2V communication. The agency's primary concerns were technical feasibility, privacy/security, estimates of cost and safety benefits. Like driverless vehicles, communication standards will take some time to gain consumer trust, but eventually they will make their way into the mainstream to compliment the progress of this industry. [Rules for Self-Driving Cars in Legal Gray Area ]
Below are excerpts from IEEE's related roundtable discussion, which you can watch in full in this video.
Jeffrey Miller: The world as we know it now is driven by wireless technologies. Most people have cellphones, we have wireless internet connections, and there are a lot of different technologies that are in use there. Having vehicles that are able to talk to each other or are able to communicate with the road way is nothing new, it's the next logical progression that we have.There are cars already that act as hotspots and communicate on the cellular network and they provide Internet access to the passengers of the vehicle. So vehicle-to-vehicle technology is just allowing two vehicles that are within proximity of each other the ability to communicate. This is something that is not difficult to do, we have short range personal networks like Bluetooth, we have dedicated short range communication, there's even cellular providers who are saying that we don't need to have the vehicles with each other over an add-hock network but perhaps they still communicate through the infrastructure and when they hit on of the base stations it comes back to communicate with one of the vehicles that are in close proximity to it. So the technology seems like it's there, we're going to need to increase the bandwidth and we are constantly improving the networks that we have but that's some of the technologies that are used for V2V.
Justin Prichard: I'd like to shift gears a little bit here, one of the other aspects of these technological advances, is vehicle-to-infrastructure communication. In other words, a car might talk to a sensor on a stoplight or a sign on the side of the road. I'd like you to talk about where we are with that.
Wei-Bin Zhang: As we were talking, most of the car manufactures now are talking about autonomous vehicles — meaning that they place intelligence on to the vehicle and have that vehicle detect everything a driver does and be able to react like a driver does. When we design an automated system, even autonomous vehicle, it does work with infrastructure, it works with land markings, it works with signs and signals and so forth. If we take a step back and take a look to say do we need to have the autonomous vehicle to totally duplicate the driver because we know that there are some limitations of drivers and the car is currently designed to somehow coop with those limitation. Automated vehicles could potentially overcome those. Giving you some examples, you already mentioned signals that are talking to cars and knowing when a signal is going to change, now to actually see the change. But also it can potentially place some infrastructure sensors, you can detect the places where the typical sensor with line of sight limitations would not be able to see. There are a lot of things that can be done. Even you could potentially build up the physical infrastructures allowing some of these problems to be isolated — other vehicles intruding or so forth. It is very important, specifically here with vehicle-to-infrastructure communication, where it is being looked at by the community currently and the USDOT, for example is leading the effort to do the connected vehicles and began making an effort earlier to this year to define, basically, what is the role of vehicle-to-infrastructure communication?
Follow all of the Expert Voices issues and debates — and become part of the discussion — on Facebook, Twitter and Google+. The views expressed are those of the author and do not necessarily reflect the views of the publisher. This version of the article was originally published on Live Science.
