Imagine cars driving themselves seamlessly and safely down the highway. Gone are traffic jams, crazy drivers, and commute-induced headaches. A vision of the future? Actually, no, this was the vision of the 1939 World's Fair and an idea that has come, come again, and now seems closer than ever before. A combination of technological availability, continuing safety concerns, and escalating congestion and energy problems fuel transportation researcher and policy-maker dreams of robotic vehicles.
Robotic or autonomous vehicles use a combination of computer vision, radar, GPS, and LIDAR to navigate the road and traffic. Many of these technologies are already being integrated into today's vehicles. If you recently purchased a new car, chances are it contains "active safety" systems that actively detect and mitigate a possible accident. For example, your car's brakes may pre-pressurize to reduce the crash severity if sensors detect an imminent crash you will be unable to avoid. Beyond safety, elements of autonomous systems offer convenience. If your car has an auto-park system, you can take your hands from the wheel and let your vehicle automatically park in that tight but precious parking spot on a busy city street.
While robot cars have always been a rich source of material for science fiction (remember Arnold Schwarzenegger in Total Recall jumping into a robotic Johnny Cab?), autonomous vehicles have been in books and research labs for several decades. Building upon work funded by the Department of Defense and conducted by industry as well as leading researchers at MIT (see the MIT CSAIL video), Stanford, Carnegie Mellon, and other universities, the Google Car is the most recent platform to provide both a symbol and demonstration of the possibility of automation to revolutionize surface transportation.
A robot-driven highway system would allay safety fears caused by badly behaved operators that cause accidents out of rage, substance abuse, or distraction. Given the coming wave of older baby-boomer drivers who may choose to give up driving or may lose their capacity to drive safely, a "driverless car" could provide unlimited on-demand mobility and freedom across the lifespan.
Autonomous vehicles would enable intelligent management of the nation's highways, optimizing traffic speed and the surface transportation system's energy consumption. Given the growing commute time for individual drivers and the billions of dollars lost to congestion-related delays, any alternative that reduces aggravation and time on the road is a win for everyone. Moreover, less idle time means less burning of whatever the fuel of the future is likely to be. So, given the decades of dreams, technological advances, and the possible benefits, what will stop these cars from rolling down the road?
Even after the considerable technological development is complete, it is unlikely that any of us will be jumping into our robotic roadster anytime soon. The technological barriers are likely to pale in contrast to the policy, market, and consumer challenges to the driverless car. Here is a partial list of issues and questions that should be on government, industry, and researcher agendas today if we are to ever realize the promise of the driverless car tomorrow:
New infrastructure: Who will blink first: the government or vehicle manufacturers? If the slow move toward determining who pays, builds, and manages an electric vehicle or natural gas vehicle infrastructure is any indication of the speed with which public and private institutions can reach consensus, imagine how long it may take to adapt today's driver-centered system to tomorrow's autonomous system? How will a highway system in transition, shared between robots and humans, be operated? (What would the road signs look like for a road accommodating passenger cars, commercial traffic, bicycles, and robots?) Who will manage it? Will the existing, primarily private, infrastructure of vehicle fueling and maintenance be converted or see another set of providers emerge?
Management and operations: Autonomous systems are likely to rely on some shared resources, cloud-based or otherwise. Who will manage and pay for this enormous infrastructure? If government or private companies, will the system be regional, national, or individual to a defined road system (e.g., "Robot Route 66")? Just consider federal government concerns for the security of these cyber-physical systems when we begin to imagine hundreds of millions of people essentially riding on the grid of a shared system (essentially a single virtual vehicle) during "rush" hour across the nation.
Consumer learning, trust, and acceptance: Today's vehicles are already including advanced systems that take an active role in the driving task. Collision avoidance systems, intelligent cruise control, various warning systems, and driver fatigue detection are a few options already on the showroom floor. Dealers are discovering that even these incremental innovations require buyer education. (Remember how long it took for people to understand and use antilock brakes correctly?) MIT Center for Transportation & Logistics and MIT AgeLab research shows that consumers of all ages need more than a product description and price list to learn, trust, and ultimately adopt new technologies as the car transitions from a system of human-in-the-loop to human-in-the-passenger-seat.
Insurance and liability: When made commercially available, robotic cars are likely to be 99.9 percent reliable, but all systems have their problems. A car that becomes a computer on wheels is likely to have some of the same problems as the computer on your desk. Who will be responsible for system safety failures? Most transportation accidents across all modes are found to be the fault of the operator -- especially when that operator is dead. Who do we blame when there is no driver? The manufacturer? The computer developer? The software code writer? The public, private, or public-private organization that is managing the highway system? What does this mean for insurers who underwrite and sell automobile insurance?
Commercial implications: Car buying has rarely been about transportation. Automobile manufacturers design dreams on wheels that promise sex appeal, freedom, independence, and the joy of power and control behind the wheel. If our car becomes autonomous, are we buying an experience or simply a ride? If it's just a ride, then it becomes basically a service, so do automobile manufacturers and their dealers become utilities or service providers instead of distributors of branded design and engineering? If so, how much is that worth to the consumer? Will the driverless car remain the second most expensive asset after our homes, like the car is today?
Policy almost always lags behind technological development. Nearly every manufacturer has plans to move to a fully automated vehicle. GM, for example, plans to integrate autonomous technology in some of its cars by 2015 and forecasts fully automated systems in little more than a decade. The U.S. Department of Transportation has been working to address many of these issues, but transportation policy and ultimately the integration of sensors, pavement, and partnerships demand the coordination of multiple levels of government spanning years well beyond the tenure of almost all elected officials.
Earlier this year Nevada was the first state to register a driverless car defined as "a motor vehicle that uses artificial intelligence, sensors and global positioning system coordinates to drive itself without the active intervention of a human operator." Other states are considering similar legislation. While the technology may be coming and some government agencies may be willing, the realization of a robotic transportation future will require a coordinated and comprehensive discussion between all levels of government and multiple industries that must begin today if we are to translate technological advances into policy and market innovation. Equally urgent will be the engineering of new dreams of what driving is for tomorrow's driver -- or should I say "passenger"?