The Autonomous Car – Self-Driving but Not Self-Regulating

In 1962 the creators of The Jetsons imagined the car of the future: it could fly, change shape and even fold into a briefcase. But it didn’t drive itself. Now, in 2016, self-driving cars—and trucks—have arrived.

The Uber-owned start-up company, Otto, has just made its first run with a driverless truck delivering Budweiser beer. Google, General Motors, Tesla, Ford, BMW and others are all looking at ways to reduce or eliminate the input of human drivers.

The president of Lyft, another San Francisco-based rideshare company, has predicted, perhaps optimistically, that in five years the majority of rides using its on-demand network will be via autonomous cars. At first autonomous cars will have restricted use and require human monitoring and intervention, but as the technology improves they will be available in more and more situations. For example, Uber has started testing its self-driving cars in Pittsburgh.

While autonomous cars will simplify public life, the regulation of them is likely to be challenging and complex, in part because the technology is evolving so quickly. Some regulation already exists. From 2011 to 2016 several U.S. states, including California, passed legislation related to autonomous cars. In September, 2014, California’s regulations for testing autonomous vehicles came into effect. Just over a year later, in December, 2015, California’s Department of Motor Vehicles released draft regulations for the deployment of autonomous vehicles for public use and operation.

In September, 2015, Ontario was the first Canadian jurisdiction to follow California and produce regulations for the testing of autonomous vehicles. We understand that the British Columbia government is monitoring developments and tests being carried out in other jurisdictions. However, B.C. appears to be in a regulatory ‘wait and see’ mode: the Motor Vehicle Act remains silent on the subject of self-driving cars, perhaps to see how effective different regulatory models have been in other jurisdictions.

Ontario provides an example of a regulatory model for the introduction of these vehicles. The Ontario regulation, Pilot Project – Automated Vehicles (306/15), created under the Ontario Highway Traffic Act, came into force as of January 1, 2016, and expires on January 1, 2026. It is intended to apply only for the purpose of testing autonomous cars on Ontario roads and so, like California’s 2014 regulations, can only be used for those limited purposes.

The Ontario regulation applies to cars equipped with an automated driving system that operates with conditional, high or full autonomy. The regulation excludes car that are equipped with ‘driver’s assistance’ and are only partially autonomous. These cars have an automated driving assistant that warns drivers of approaching pedestrians, and the car may even automatically break in an emergency. The driving assistant also detects lane markings and issues a warning to the driver by vibrating the steering wheel whenever there is a threat of leaving the lane unintentionally at speeds above approximately 70 km/h. Cars that do these limited driving tasks but are not fully automated will be governed by the same legislation as any other vehicle in Ontario.

In contrast, the 2015 California draft regulations anticipate public use of autonomous cars and do not attempt to classify cars by the level of autonomy of the operating system. They define autonomous cars as ‘any vehicle equipped with technology that has the capability of operating or driving without the active physical control or monitoring of a natural person, whether or not the technology is engaged’.

Both the Ontario and California regulations have very restrictive provisions. The individual who sits in the driver’s seat is considered to be the driver of the vehicle, even if the autonomous driving system is engaged. That driver must still hold a valid driver’s licence, and is required to remain in the driver’s seat at all times and monitor the vehicle’s operation. In the early stages of implementation of self-driving technology, this might be characterized more as ‘mutual cooperation’ than autonomous control. In Ontario, these regulations also require that if the autonomous system fails and the driver is unable to take over, then the vehicle must be capable of removing itself safely from traffic.

Drivers will be very interested in the question of who will be responsible for traffic violations when the driver is merely monitoring an autonomous car. Ontario’s regulations do not explicitly state that the driver of the vehicle will be responsible for all traffic violations. They require only that in the event of a collision or traffic stop, the driver of the automated vehicle shall advise the attending police officer that he or she is operating an automated vehicle and it is being tested under the pilot project. In contrast, California’s draft regulations for the public use of autonomous cars provide that the driver of an autonomous car is responsible for all traffic violations, as it is anticipated that autonomous cars will make mistakes and the drivers must take responsibility.

Lawyers, in turn, will be very interested in the liability issues that will arise if and when the autonomous technology fails.

It is expected that the use of autonomous cars will significantly reduce motor vehicle accidents. There are an estimated 30,000 deaths in motor vehicle accidents in the U.S. every year, and a significantly higher number of injuries. Studies indicate that more than 90% of motor vehicle crashes are caused at least in part by human error. Significantly reducing those numbers may result in substantial savings in both health care and legal costs to governments and insurers such that governments may eventually deem it worthwhile to support autonomous car technology.

In the aftermath of a fatal crash in Germany which involved an autonomous car, German lawmakers proposed mandatory ‘black boxes’ for self-driving cars to record data regarding how they are being controlled in the moments before an accident. Black boxes are already being installed in most new vehicles, and they record data including information about the speed of the car and who was or was not wearing their seatbelt.

A reliable source of such information may transform the way our justice system resolves issues of liability after a motor vehicle accident. It may also allow companies that create autonomous driving technologies to mount a proper defence in the event of an accident involving their products. As well, the data will likely be able to indicate to regulators and manufactures what ‘mistakes’ autonomous vehicles continue to make. This will help manufacturers improve their products and assist regulators in setting safety standards for the vehicles.

Fully driverless cars remain ‘the car of the future’ and, perhaps, a solution for other problems. They may eventually offer newfound autonomy and flexibility to those who, due to age, disability or other reasons, are restricted from driving. But for the immediately foreseeable future, they will remain strictly regulated.

For more information on self-driving vehicles and their legal implications, please contact David.

For more information, please contact: