Tesla could be working on gaze-based driver attention tracking

Although the company has insisted that it isn't meant to be used for full self-driving purposes, Tesla's Autopilot feature has still be presumed by some drivers to mean they can take their eyes off the road even while still holding the wheel. Accidents involving Tesla's cars have often blamed this feature and the company might be looking into ways to curb that before the numbers rise. Tesla EVs might soon be alerting drivers if their attentions veer off course thanks to hardware that's already present in some models.

The Tesla Model 3 has had a cabin-facing camera for years but it was only last year that Elon Musk revealed what it would be for, along with the assurance that the cameras have been inactive all this time. Although incapable of being fully self-driving, the cameras are already preparing for future Tesla robotaxis where passenger behavior would be recorded for potential legal purposes.

Earlier this year, though, Tesla activated the cameras in the Model 3 and Model Y cars on a voluntary basis to collect images and short video clips of the cabin prior to a collision. Of course, drivers were free to agree or opt-out of the study as they wish, but one popular Tesla hacker has uncovered more than what Tesla may have said it was collecting.

Green revealed that Tesla's software was also detecting events like when the driver's eyes are closed, up, or down or whether the driver is using the phone. These are pretty much key data points for detecting driver attention, something that Tesla hasn't hinted it was even looking into. Tesla's Autopilot currently only detects the torque that's being applied to the steering wheel.

Tesla wouldn't be the first car maker to use such cues to measure driver attention. Whether it will be effective, however, depends on whether the carmaker actually implements such a feature, something it has been rather silent about despite the many accidents involving its Autopilot feature.