Two years ago, the National Highway Traffic Safety Administration (NHTSA) announced that after Tesla introduced their Autopilot mode, crash rates for Tesla brand cars dropped by 40%.
When R.A. Whitfield saw those numbers, he couldn’t believe it. Whitfield, the director of Quality Control Systems Corp. and an expert in statistics, requested the data from the NHTSA. He encountered a wall of bureaucracy and engaged in a two-year lawsuit that finally uncovered the data in November 2018.
In a 25-page response that Whitfield published on February 8, he details his discoveries of the NHTSA’s violation of basic principles of research methodology as well as describing that the data draws no conclusion whatsoever.
In response, the NHTSA did not dispute Whitfield’s report, but did say that they would asses the Quality Control System Corp. report and “provide comment as necessary.”
“The lesson here is a need for candor,” said Bryant Walker Smith, a driverless-vehicle law expert at the University of South Carolina. “What does it mean to be trustworthy in this field?”
Smith wants to make sure companies, regulatory agencies and politicians communicate clearly. “This is what we’re doing, this is why we think it’s safe, and this is why you should believe us.”
While the field of driverless cars is advancing rapidly, the law has been slow to catch up.
At the moment, federal law does little to regulate autonomous vehicles. Instead, the burden is on states, who have a myriad of resolutions with no nationwide consistency.
Congress tried to pass a federal bill last year, but the proposed bill failed in the senate. Smith attributes this to a lack of trust in the technology. A recent Gallup survey showed that 52% of those polled never wanted to use a self-driving car.
The main problem Whitfield identified was that the NHTSA took air bag deployments before and after Autosteer installation to make and estimate of the number of crashes per million miles. But most of the cars reported by Tesla were missing the miles the car traveled before Autosteer was installed. With no miles at all to add to the equation, but the same number of air bag deployments, Whitfield claims that any findings would inflate the crash rate for pre-Autosteer cars.
He actually found that when mileage data was correctly recorded, Tesla cars were involved in 60% more crashes. This means that Autosteer could be even more dangerous than a human driver.
“We did not produce this data, and don’t vouch for it,” he said. “We don’t know if it’s true or not.”
Nonetheless, Whitfield refused to draw any conclusions and maintains that he published his report in the name of public safety. “Efforts to hide the crash record will impede progress in achieving whatever safety beneﬁts advanced driver-assistance systems might ultimately bring.”