Vancouver City (the “City”) Councillor Geoff Meggs plans to table a council meeting motion this Tuesday to have the City investigate putting self-driving cars onto Vancouver roads:
http://vancouver.24hrs.ca/2016/02/18/speed-up-on-driverless-cars-meggs
However, the Province of British Columbia (the “Province”) Transportation Minister Todd Stone has stated that the City does not have jurisdiction to create a pilot project to test self-driving cars:
http://www.cbc.ca/news/canada/british-columbia/driverless-cars-bc-vancouver-1.3454762
In the event that self-driving cars do become a reality, it is likely that the Province and the City will put aside their differences and create a comprehensive set of laws and regulations (and likely accompanying fees) to accommodate the new technology, but how close is this technology to being a day to day reality? Various companies have participated in pilot projects to develop and test self-driving cars, including: Google, Tesla, Audi, BMW, Honda, and Delphi. There are conflicting reports about just how safe self-driving cars are.
In June of 2015, Google stated that their self-driving cars, after over 6 years and 1.8 million miles of driving, had been involved in only 12 accidents, all of which were minor and caused by the human drivers of other vehicles:
However, a closer look at the data reveals that this may not be the whole story:
Google has been forced to make hundreds of human interventions to prevent accidents. Other factors, such as the type of conditions in which these cars are driving in, may be important factors as well. For example, if the self-driving cars are logging their miles on familiar roads, in lighter weather conditions, or in low traffic areas, it would wildly skew data. Overall, it’s much easier to manufacture positive safety data using controlled conditions. A recent study out of the University of Michigan has found that, despite these less demanding testing conditions, self-driving cars are twice as likely to be involved in motor vehicle accidents:
http://time.com/4098303/self-driving-cars-accident/
Overall, many of the issues seem to arise from poor interaction between human drivers and self-driving vehicles:
Humans often bend rules to avoid accidents, and their driving behaviours aren’t as calculated as their automated self-driving counterparts. This suggests that it may not be possible to slowly introduce self-driving cars onto the roads. Instead, we may have to wait until a large swath of self-driving cars can be introduced at once or existing vehicles can cheaply be altered to become self-driving.
Putting current safety concerns aside, it is likely that self-driving cars will become a reality at some point, as the technology matures. This creates serious concerns for people injured by self-driving cars. Most notably, who is legally responsible when they malfunction or a dispute over liability for an accident occurs. Currently, once the identity of a driver has been ascertained, an injured party can then sue that driver. A driver-less vehicle obviously has no driver, and suing the large corporate manufacturer of the vehicle is surely to be a daunting task.
