Tesla Believes Autopilot Update Would Have Saved Driver's Life In Fatal Crash

Photo credit: Spencer Platt/Getty Images
Photo credit: Spencer Platt/Getty Images

A Tesla Model S driver became the first human to die in a vehicle driving semi-autonomously this past May when it struck an 18-wheeler with its Autopilot system enabled. In a press conference on Sunday, Tesla CEO Elon Musk said the company believes its new Autopilot update would have saved the driver’s life.


Tesla blamed auto-braking system for the crash in late July, saying the it was not a failure of Autopilot. But the bottom line is that, in Tesla’s own words, “neither the driver nor the car’s sensors could see a tractor trailer pulling across the highway.” But had the car been equipped with the update to the Autopilot system, which primarily depends on radar instead of camera detection, Musk said it probably would have saved Brown’s life.

“We can’t say it with absolute certainty, but we believe that yes it would have [saved Brown’s life],” Musk said. “It would see a large metal object across the road, and there’s no road sign there. This would not be a whitelisted situation, so it would brake.”

The original Tesla Autopilot system used cameras as its primary method of feeding information about the surrounding environment to the car. Camera sensors kept the car in its lane, and mostly prevented it from making contact with other objects. But as a Model S owner demonstrated in a YouTube video, the calls got uncomfortably close on occasion.

Two months after that video published on YouTube, Ohio resident Joshua Brown collided with an 18-wheeler in Florida with his Model S on Autopilot. Potentially consistent with a common misunderstanding that driver-assistance systems like Autopilot are capable of driving without attention or intervention from the driver—despite cars displaying a warning contradicting that upon being turned on—there was belief that Brown was watching a Harry Potter movie at the time of the crash. Authorities are still unsure as to if that is true or not.


When asked if the update was in response to Brown’s crash, Musk said it’s something he’s been wanting to do since late last year. Tesla said it added radar capabilities to all of its cars in October 2014 as part of the Autopilot hardware suite, so the capability to make this change has been there for nearly two years.

But Musk called solving the problem of radar use was a hard one, as a Tesla blog post described the goal of the system as “creat[ing] a picture of the world using the onboard radar.” That’s pretty familiar to those of us who know of LiDAR detection and the like, but Musk said “it’s simply not possible” to make this approach work without testing on fleet-connected cars.


“I was always told, ‘No it’s not possible, you can’t do it, it’s not going to work, nobody else has made it work, software’s too hard, sensor’s not good enough,’” Musk said. “I really pushed hard on questioning those assumptions in the last three or four months, and now we believe that there is [a way to make it work].”

With the new update, the Tesla blog post said the car should almost always hit brakes correctly—“even if a UFO were to land on the freeway in zero visibility conditions.” Here’s more from the post:

After careful consideration, we now believe [radar] can be used as a primary control sensor without requiring the camera to confirm visual image recognition. This is a non-trivial and counter-intuitive problem, because of how strange the world looks in radar. Photons of that wavelength travel easy through fog, dust, rain and snow, but anything metallic looks like a mirror. The radar can see people, but they appear partially translucent. Something made of wood or painted plastic, though opaque to a person, is almost as transparent as glass to radar.

On the other hand, any metal surface with a dish shape is not only reflective, but also amplifies the reflected signal to many times its actual size. A discarded soda can on the road, with its concave bottom facing towards you can appear to be a large and dangerous obstacle, but you would definitely not want to slam on the brakes to avoid it.

Therefore, the big problem in using radar to stop the car is avoiding false alarms. Slamming on the brakes is critical if you are about to hit something large and solid, but not if you are merely about to run over a soda can. Having lots of unnecessary braking events would at best be very annoying and at worst cause injury.


The blog post itself has the exact steps Tesla took to weed out the problems associated by leaning on radar as the primary control sensor, including how it detects road signs and assesses the movement of other objects with the radar.

Tesla said that because of the radar, the car will be able to see through rain, fog, snow and dust easily. But with fully autonomous cars still likely a few years away, Autopilot remains an assistance system. Regardless of the situation, the driver must stay alert and attentive to the road and his or her surroundings.


farscythe - makin da cawfee!

just... stop calling it autopilot untill it can drive you from a to b without your help..


(also saying this new software would have saved the guy(probably) is like saying seatbelts would have saved the guy in the 50s).. it doesnt strike me as a very smart/sensitive thing to say