How well does Tesla’s autopilot detect lane-splitting motorcycles?
A Tesla owner’s experience shows there’s still a ways to go.
This story was originally published on motorcyclistonline.com.
Lane-splitting is a benefit to riders all over the world, but in the United States, the privilege is currently reserved to riders in the Golden State. California-based Tesla seems to know a thing or two about motorcyclist’s unique advantage in the state, and by all appearances the company is working to ensure that riders filtering through traffic stay safe when its autopilot systems are active. But does it work? Can Tesla’s Autopilot detect a lane-sharing rider?
Sort of. We can see Tesla’s Autopilot in action thanks to YouTuber Scott Kubo, who recently put together the video below to test the functionality of motorcycle detection while lane-splitting.
Self-driving cars are still a burgeoning field of technology, and it’s apparent there are still bugs to work out.
Detecting an approaching motorcycle seems to be hit and miss in Version 9 of Tesla’s neural net Autopilot software. It’s clear that the system confuses motorcycles with cars at times, and can even miss a motorcycle entirely if the motorcycle is moving at a good clip. It’s tough to determine how useful this detection would be in its current state. Being able to separate a lane-sharing motorcycle is critical when stopping or correcting an in-progress lane change, for example. It’s a start though—and an admirable one at that.
Teslas are well represented on California highways, and more manufacturers seem to take up the autonomous driving challenge every year. Any technology that helps keep riders safe on the road—wherever that road might be—is a positive one in my book.
Check out the video and tell us what you think. Where should technology like this go to be truly useful to riders?