HAV or HAV-not: DOT ponders a self-driving future

HAV stands for “highly automated vehicle.” Regulators around the world are struggling to keep up with what they believe — or have been convinced of — is the fast-coming advent of HAVs. Trust me, you can get lost in this, but the idea of highly automated runs from a car that takes over from the drive in a danger situation to autonomous — you sit in a bubble with lounge chairs and it drives itself to where you want to go.

If the early automobile replaced the horse,  autonomy promises autos to replace the jackass behind the wheel.

Autonomy didn’t start with Google. Something called the Prometheus European Project spawned a Mercedes that could zoom along the autobahn all by itself back in 1994. But these things have a way of percolating for a long time. Now an ecosystem incorporating Silicon Valley, the auto industry, and would-be commercial users of self-driving vehicles has sprung up, so the idea is getting a lot of push.

Just as drones have the FAA tied up in knots over how to regulate them, self-driving cars vex the Transportation Department’s National Highway Traffic Safety Administration. Both efforts share the challenge of how a disruptive technology collides with established norms with people at the controls.

Given the potential mayhem driverless cars and drones could cause, this seems like a logical area for a federal regulatory framework, lest a Chevro-Google-Let in San Francisco operate as an outlaw in Dallas.

Last month’s release of the 116-page Federal Automated Vehicles Policy came as states and some large cities issue their own regulations for driverless cars — in many instances because of demand from commercial entities. Another case of software replacing meatware. In smaller cities, such as Ann Arbor, Michigan and Pittsburgh, Domino’s Pizza and Uber, respectively, are testing this. On the other hand, Chicago is considering ordinances banning self-driving cars.

The DOT policy isn’t vehicle regulation. That is yet to come. But it has produced controversy. One critic, Paul Brubaker, praises NHTSA’s effort but says it falls short for being derived from a 1960s era regulatory framework — the one that gave us the seatbelt mandate and the PRND transmission rule. Brubaker is CEO of the nonprofit Alliance for Transportation Innovation (ATI21). Brubaker has been around Washington for a long time. He was administrator of the Research and Innovative Technology Administration at DOT for a spell. Alliance members consist of a mixture of software companies, other associations and transportation authorities.

In my Federal Drive interview with him, he says the “innovators” were absent from the comment period leading to the policy. He says NHTSA administrator Mark Rosekind is aware of this and wants more input. ATI21 just published its own analog to the DOT policy. It proposes ways everyone affected by a driverless future can get together on a more modern policy framework.

ATI21 and the Transportation Department agree that safety from cars that, in theory, don’t crash, tops the list of drivers of a driverless automotive future. Basically every accident is caused by driver error, except for incidents like locked-up ignition switches and exploding air bags. But cutting the cost of car ownership, better access to transportation for the disabled, and a more energy-efficiency add to the case. Crashless presupposed software that’s better than human judgment. The technology isn’t perfected yet, but it’s getting close.

I wonder how autonomy would figure in extraordinary situations. In 1962, French President Charles De Gaulle escaped an assassination attempt when his driver zoomed out of a skid in his Citroen DS-19, all tires flat and windows shot out with bullets. Think too of when President Ronald Reagan’s Lincoln limousine sped him to GW Hospital, diverting after initially heading to the White House. How do you program that?

Comments