When we think of driverless (or autonomous) cars, we often picture a futuristic time in which someone is relaxing in the passenger seat while the vehicle safely navigates its way to the desired destination. Believe it or not, certain aspects of that kind of technology already exist. Toyota Safety Sense, Active Cruise Control from BMW, Ford’s AdvanceTrac Electronic Stability Control, Volkswagen’s Park Assist, and Tesla’s Enhanced Autopilot are all examples of “smart” technology that enables a car to take certain actions so its driver doesn’t have to. Most self-driving technology currently in place is designed to make driving safer.
But how far are we from seeing fully autonomous cars on the road? According to the RAND Corporation (a research organization that, in response to public policy challenges, develops solutions to help make communities throughout the world safer, more secure, healthier, and more prosperous), “Autonomous vehicles would have to be driven hundreds of millions of miles and, under some scenarios, hundreds of billions of miles, to create enough data to clearly demonstrate their safety.”
Yet, some experts, such as James Kuffner, chief technology officer of the Toyota Research Institute, believe we must avoid using the number of miles driven as a metric of capability. He said driving during rush hour is different from driving on an open road in South Dakota. Complexity is key.
Since 2009, Google (now Waymo) has driven 3 million miles in autonomous mode, in which software drives the vehicle. Although the company has stopped publishing monthly accident reports for its self-driving cars, it has had very few accidents throughout the course of its testing. In fact, according to the last Google Self-Driving Car Project Monthly Report (November 2016), the company had no traffic collisions to report.
In a 2016 episode of Bloomberg West, John Leonard, a computer science and A.I. lab professor at MIT, said he believes cars will become highly automated in just five years. Daniela Rus, also a computer science and A.I. lab professor at MIT, said we already have self-driving cars that are capable of driving at low speeds and in low-complexity environments. The question is: how can we scale self-driving vehicles so that we can drive them in more challenging and populated environments? Because of these constraints, Richard Wallace, director of transportation systems analysis for the Center for Automotive Research, believes the technology to make it work will be developed before regulations are in place.
Nonetheless, all industry leaders concur that today we are not in a place where we can trust that vehicles can drive us safely from point A to point B without our participation, especially in uncontrolled environments and scenarios.
In a new guide, we explore the industry’s interest in and movement toward autonomous and connected cars. You can download it here.