National Highway Traffic Safety Administration Archives

The Driverless Commute: GM’s driverless exemption petition advances, but key questions for industry remains; a driver’s test for driverless cars; and robo-race car to edge test

Posted on March 29, 2019 By Eric Tanenblatt and James Richardson
Categories: Autonomous Vehicles, Driverless Commute

1. NHTSA advances GM’s FMVSS exemption bid. But we still don’t have answers on liability.

Car from GM Cruise LLC, a driverless car company that tests and develops autonomous car technology.

The National Highway Traffic Safety Administration is airing for public comment two petitions to deploy on public roads vehicles that lack conventional controls like a steering wheel or pedals.

The move came some fourteen months after General Motors first asked federal regulators for a temporary exemption from Federal Motor Vehicle Safety Standards. The automaker asked the agency in January 2018 for 16 human-driver-based exemptions from FMVSS with the hope it could deploy a fleet of robocabs later this year. (You can read their petition here.)

If approved, federal regulators would be endorsing the bold proposition that autonomous vehicles (these ones, at least) can deliver a standard of safety equivalent to what is already required of existing cars, but the long wait is evidence that catching the feds’ green light is neither easy nor assured.

The Driverless Commute, presented by Dentons: How programmer bias will be reflected in AI platforms; a case for public investment in AV access; and Waymo, GM & Ford top new AV leaderboard

Posted on March 13, 2019 By Eric Tanenblatt and James Richardson
Categories: Autonomous Vehicles, Driverless Commute

1. When an engineer’s implicit bias becomes a computer’s

Can the machine learning algorithms that underpin the operation of autonomous vehicles perpetuate—or worsen, even—social, structural biases against people of color?

Maybe, according to a new paper (PDF) from the Georgia Institute of Technology, where researchers tested the accuracy of object detection systems (not unlike those used in driverless cars) in positively identifying pedestrians of varying skin colors.

Researchers paid human test subjects to review a collection of 3,500 images of people of varying skin tones and to mark each photograph with “LS” or “DS” to designate light or dark skin of the subject.