Tesla unveils updates to Autopilot feature

Tesla unveils updates to Autopilot feature
© Getty Images

Tesla Motors unveiled updates to its Autopilot feature after the semi-autonomous technology was involved in a fatal crash earlier this year.

The automaker announced in a blog post that it would be upgrading the Autopilot software to primarily rely on radar sensors instead of cameras, which will help the vehicle paint a better picture of its surroundings.


All Tesla vehicles had been equipped with radar sensors since October 2014, but it “was only meant to be a supplementary sensor to the primary camera and image processing system.”

“After careful consideration, we now believe it can be used as a primary control sensor without requiring the camera to confirm visual image recognition,” Tesla said.

The announcement comes after several crashes involving cars that were engaged in Autopilot, including one deadly incident in which a vehicle slammed into a white tractor-trailer after not being able to differentiate the truck from the brightly lit sky around it.

Tesla has been facing mounting criticism for whether the feature lulls drivers into a false sense of security and whether the company rolled out the technology — which is still in the public beta-testing phase — too quickly.

Drivers are still required to be behind the wheel while using Autopilot, but the car can change lanes, brake, steer, accelerate, decelerate and avoid obstacles on its own.

In the new upgrade, “Autosteer” will repeatedly warn drivers to keep their hands on the steering wheel. If the warnings are ignored, drivers will have to park the car in order to restart the feature.

Tesla explained that the shift to radar can seem counterintuitive because of how “strange the world looks in radar.” A soda can on the road, for example, can look like a large, dangerous obstacle in radar.

“Therefore, the big problem in using radar to stop the car is avoiding false alarms,” the company said. “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.”

Tesla said the new software upgrade will have a more detailed data point cloud and take radar snapshots every tenth of a second in order to assemble a 3D picture.

Fleet learning also will help improve the technology as more vehicles upload travel information to a database, Tesla said.

The software upgrades are slated to take effect in two weeks.