Autonomous Vehicles: Feds Step on the Data-Sharing Pedal
Since publishing our series on autonomous vehicles, there's much to report, including a new policy highlighting data sharing to improve safety.
- By Barry Devlin
- October 5, 2016
In little more than two months since I published an in-depth review of the implications of autonomous vehicles for IT, employment, and society at large, the topic has switched to the fast lane of public interest. This week, even the US government has taken to the road to promote sharing of autonomous vehicle data… to perhaps predictable industry reactions. But first, here's the rest of the news.
Just as the above series was going to press, the first fatality involving a semi-autonomous vehicle was being reported. The accident, which actually occurred in May, involved a Tesla Model S crashing into the side of a truck that was turning left across its path. Much has been said about the ethical choices that autonomous vehicles may have to make, such as choosing between protecting the life of driver or that of a group of school children crossing the road. We are still far from that dilemma. Tesla stated that "[n]either autopilot nor the driver noticed the white side of the tractor-trailer against a brightly lit sky, so the brake was not applied."
Previous speculation that the first fatality in a self-driving vehicle would raise a storm of concern among a frightened public has so far proven baseless. Uber offered its first self-driving rides to the general public on the streets of Pittsburg on September 14th. No panic on the streets ensued. Of course, the cars were also equipped with human drivers who could take over from the autopilot in the event of any problem -- and to comply with current law. The 2014 vision of Uber CEO, Travis Kalanick, of replacing all his drivers with autonomous vehicles clearly has some way to go, but few were predicting even this experimental public roll-out so soon.
Not to be outdone, Lyft co-founder, John Zimmer, took to the Internet a few days later to proclaim that "[a]utonomous vehicle[s] ... will account for the majority of Lyft rides within 5 years." Perhaps more interesting, he opined on a topic I covered in Part 4 of my series: the effect of vastly reduced vehicle numbers on our cities. Predicting that private car ownership will be a rarity in major U.S. cities by 2025, Zimmer suggests that we have a once-in-a-century opportunity to redefine urban living in a way that benefits everybody and improves the lives of the growing urban population by returning the streets to people.
On the employment front, further analysis of the impact of autonomous vehicles by Wolf Richter in a September 14th blog suggests that the potential number of jobs at risk in the U.S. could be as large as 4 million. Given that drivers constitute a significant financial cost to industries from trucking to taxis and ridesharing, Richter sees the trend toward autonomous vehicles as unstoppable. I agree.
Returning to the topic of autonomous vehicle data, on September 20, the U.S. Department of Transportation's National Highway Traffic Safety Administration (NHTSA) released a 112-page Federal Automated Vehicles Policy. It focuses on enabling "the safe deployment of automated technologies because of the enormous promise they hold to address the overwhelming majority of crashes and save lives" according to NHTSA administrator Dr. Mark Rosekind. With 94 percent of U.S. crashes due to human choice or error, the administration believes that autonomous vehicles have the potential to save thousands of lives.
Recording and sharing of data between vehicles, manufacturers, and government agencies is featured prominently. With a focus on safety, the emphasis in the policy document is on data such as that related to failures, accidents, security hacks so analysis and reconstruction of any such events are possible. In a rapidly evolving and competitive market, manufacturers' reactions to such data sharing have been rather negative, despite the obvious safety benefits of shared early access to failure data.
David Strickland, spokesperson for the intriguingly named Self-Driving Coalition for Safer Streets, created by Ford and Volvo, as well as Uber, Google, and Lyft, responded by emphasizing the data and property rights of the founders, with "on the other hand, safety [as] a number one priority." Declining to comment specifically on the NHTSA new policy's requirements for data sharing, he asserted that "the devil is in the details."
Indeed. Data generated by autonomous vehicles is both extensive and comprehensive. The details embedded within it have implications far beyond safety and data ownership. Not least are the implications for privacy and personal security. Add ethical and economic considerations, and the scope of required thought leadership and regulation stretches far beyond the Department of Transportation and industry lobbyists in every jurisdiction. Who will step up to the task?
Dr. Barry Devlin defined the first data warehouse architecture in 1985 and is among the world’s foremost authorities on BI, big data, and beyond. His 2013 book, Business unIntelligence, offers a new architecture for modern information use and management.