Tesla FSD Beta 10.5 Release Notes leaked, teases significant improvements

Tesla FSD Beta 10.5 is expected to be released soon, though a copy of the update’s Release Notes has been shared online. Based on what could be seen from the upcoming update’s notes, it appears that FSD Beta 10.5 would be a significant step up from its previous iterations. 

If Tesla does release FSD Beta 10.5 this weekend, it will fall right into Elon Musk’s estimate for the update’s release. Earlier this month, Musk hinted that V10.5 would be released around November 20. Musk doubled down on this estimate on November 19, with the CEO noting on Twitter that he was hoping for FSD Beta 10.5 to be ready very soon.

It did not take long before reports emerged from the Tesla community that the latest iteration of the advanced driver-assist system was starting to roll out. Tesla update tracker service Teslascope noted in a later post on Twitter that it had been able to confirm that FSD Beta 10.5 was indeed being released, but only for employees for now. Assuming that last-minute checks were deemed satisfactory, a release to members of the FSD Beta program was reportedly still expected for Sunday night. 

Amidst these reports, the Release Notes of FSD Beta 10.5 were also shared online. Similar to previous updates, V10.5’s notes were detailed, describing exactly what was improved in the update. As could be seen from the notes, Tesla has improved FSD Beta’s VRU once more, as well as the system’s cone and sign detections. More importantly, V10.5 also provide FSD Beta with the ability to change lanes away from a merging vehicle instead of slowing down. “False slowdowns” have been reduced as well. Overall, these improvements should make FSD Beta 10.5 perform more humanlike on the road.

The following are the leaked Release Notes of FSD Beta 10.5. 

– Improved VRU (pedestrians, bicyclists, motorcycles) crossing velocity error by 20% from improved quality in our auto-labeling.

– Improved static world predictions (road lines, edges, and lane connectivity) by up to 13% using a new static world auto-labeler and adding 165K autolabeled videos.

– Improved cone and sign detections by upreving the generalized static object network with 15K more video clips and adjusting oversampling and overweighting strategies (+4.5% precision, +10.4% recall). 

– Improved cut-in detection network by 5.5% to help reduce false slowdowns.

– Enabled “emergency collision avoidance maneuvering” in shadow mode.

– Enabled behavior to lane change away from merges when safe to do so.

– Improved merge object detection recall by using multi-modal object prediction at intersections.

– Improved control for merges by increasing smoothness of arrival time constraints and considering possible merging objects beyond visibility.

– Improved lane changes by allowing larger deceleration limit in short-deadline situations.

– Improved lateral control for creeping forward to get more visibility

– Improved modeling of road boundaries on high curvature roads for finer maneuvers.

– Improved logic to stay on-route and avoid unnecessary detours/rerouting.

0 0 votes
Article Rating
Notify of
Newest Most Voted
Inline Feedbacks
View all comments
Would love your thoughts, please comment.x