Autonomous vehicle safety bill needs improvement
Last week, Pennsylvania legislators held a Pittsburgh press conference to introduce a new Highly Automated Vehicle ( HAV) regulation bill (SB 965). While the bill’s provision to permit HAV testing without human backup drivers on public streets has received the most attention, the bill would also ban local governments from making their own HAV testing regulations, driver or no driver.
Whatever benefits in investment and innovation might be promised with this bill, the cost in safety is too high. The bill is too heavily weighted toward industry interests; changes are needed to deliver on the promise of enhanced safety.
Fortunately, it is easy to promote safety, and respect the needs of other stakeholders, while still enabling economic growth and jobs.
Sen. Wayne Langerholc Jr., R-Johnstown and chairman of the PA Senate Transportation Committee, promised at the press conference that the bill would “incorporate global design standards from the Society of Automotive Engineers International [SAE] which will ensure the safe design of HAVs.” However, the bill references no actual SAE safety standard.
The Warrendale-based SAE did issue the J3016 standard mentioned in the bill, but it is a terminology standard only. It does not impose requirements on driving automation systems, and is not intended to ensure safety. I should know, because I’m a member of the SAE committee that issued it.
There is another SAE standard, however, that does cover testing safety: SAE J3018. This standard provides guidance for training safety drivers and ensuring that HAV public road testing is done in a responsible manner. Because Pittsburgh is, and will remain, a development center for HAV technology, withdriver testing will always be essential for the newest innovations. But SB 965 would make it impossible for the city to require
developers to follow this basic safety standard.
Further, SAE has issued guidance for safety driver training and the implementation of a Safety Management System that stems directly from lessons learned by Pittsburgh-based Uber ATG. That organization is no longer with us, at least in part because their unsafe testing practices resulted in a tragic fatality in Tempe, Ariz., in 2018. The hard-won economic gains of other Pittsburgh companies are similarly at risk if they fail to heed the lessons gained from Uber ATG’s costly safety mistakes.
Putting SAE J3018 into this bill would not be onerous. After all, Pittsburgh-based Argo AI already conforms to that standard, and Motional has also been evaluated for testing safety along similar lines. The New York City DOT already requires conformance to J3018, and Massachusetts is pursuing that path as well. Why would Pittsburgh companies who say that safety is
their top priority object to following their own industry standards to keep residents — and their employees’ families—– safe? This is a chance for Pittsburgh and the rest of the commonwealth to lead by example for road testing safety.
SAE J3018 conformance will require a supply of trained safety drivers to ensure that the continual stream of software updates remains safe. This is a golden opportunity for Pittsburgh to further develop its workforce beyond elite researchers to also include skilled test drivers. This approach can help provide jobs and make Pittsburgh the worldwide leader in HAV industry safety.
But the Senate bill’s incredibly broad pre-emption clause will prevent Pittsburgh, and other municipalities, from taking any action whatsoever to regulate HAVs, or to enforce state regulation of HAVs. Our hands would be tied if an unscrupulous company should decide to conduct unsafe testing here.
Pittsburgh should have the right to intervene without waiting for the commonwealth to act if unsafe testing puts its citizens at undue risk. As currently written, the bill might even be interpreted to mean that municipal police would be prohibited from stopping a recklessly driven HAV to issue a moving violation citation.
With the SAE J3018 standard and a narrower pre-emption clause, SB 965 can be a win-win, advancing safety and innovation for the good of Pittsburgh as a whole.
After all, Pittsburgh-based activities have been instrumental in creating HAV technology, and remain essential to its success. This bill is a chance to cement our long-term legacy in this area, but, in its current form, it falls far short.