Tesla FSD Beta v10.69.3 Is Rolling Out to Employees in the US & Canada

von Eva Fox November 01, 2022

Tesla FSD Beta v10.69.3 Is Rolling Out to Employees in the US & Canada

Photo: Tesla

Tesla has begun rolling out FSD Beta v10.69.3 to its employees in the US and Canada, meaning a wider release of the new version may not be far off. FSD Beta v10.69.3 received a number of major updates listed in the release notes.

Tesla has begun rolling out Full Self-Driving Beta v10.69.3 to its employees in the United States and Canada, @teslascope/Twitter reported. The new update is available to a very limited number of testers, who must thoroughly test it before it is rolled out to more program members. If any errors and bugs are noticed, then the AI ​​team will need to fix them, after which further deployment can be expected.

Elon Musk previously explained that FSD Beta v10.69.3 will be major and contain significant improvements, so its release was slightly delayed. The team had to do internal testing, which apparently lasted several weeks, before making a public release.

FSD Beta is becoming more and more reliable, so in September, the number of testers was expanded by 60% to 160,000. The manufacturer has expanded access to the software for owners with a safety score above 80, as previously promised. The move signals a surge in confidence in FSD Beta's capabilities. It seems that we will see an even wider rollout in the coming months, according to information announced during the Q3 2022 Earnings Call.

Musk said that Tesla expects FSD Beta to be released in North America as early as Q4 2022. This means that any customer who buys a Tesla car vehicle with FSD or has one and orders FSD will have access to FSD Beta. This opportunity can be provided in November. The move was made possible because the safety that the company sees when the car is in FSD is actually significantly better than the safety they see when the car is only driven by a human pilot. This is what became the key threshold for moving to a wide beta version.

“So, this quarter, we expect to go to a wide release of full self-driving Beta in North America. So, anyone who has ordered a full self-driving Beta -- full self-driving, will have access to the FSD Beta program this year, probably about a month from now. So -- and then obviously, any new -- anyone who buys a car and purchases a full self-driving option, will immediately have that available to them.”

Musk expressed confidence that Tesla will achieve full autonomy in its vehicles, stressing that the company is almost there. He also explained that after this is achieved, Tesla will have to prove its success to the regulatory authorities and gain their approval, and here the time frame is out of the manufacturer’s control.

Full Self-Driving Beta v10.69.3 Release Notes

  • Upgraded the Object Detection network to photon count video streams and retrained all parameters with the latest autolabeled datasets (with a special emphasis on low visibility scenarios).
  • Improved the architecture for better accuracy and latency, higher recall of far away vehicles, lower velocity error of crossing vehicles by 20%, and improved VRU precision by 20%.
  • Converted the VRU Velocity network to a two-stage network, which reduced latency and improved crossing pedestrian velocity error by 6%.
  • Converted the Non VRU Attributes network to a two-stage network, which reduced latency, reduced incorrect lane assignment of crossing vehicles by 45%, and reduced incorrect parked predictions by 15%.
  • Reformulated the autoregressive Vector Lanes grammar to improve precision of lanes by 9.2%, recall of lanes by 18.7%, and recall of forks by 51.1%. Includes a full network update where all components were re-trained with 3.8x the amount of data.
  • Added a new "road markings" module to the Vector Lanes neural network which improves lane topology error at intersections by 38.9%.
  • Upgraded the Occupancy Network to align with road surface instead of ego for improved detection stability and improved recall at hill crest.
  • Reduced runtime of candidate trajectory generation by approximately 80% and improved smoothness by distilling an expensive trajectory optimization procedure into a lightweight
  • Improved decision making for short deadline lane changes around gores by richer modeling of the trade-off between going off-route vs trajectory required to drive through the gore region.
  • Reduced false slowdowns for pedestrians near crosswalk by using a better model for the kinematics of the pedestrian.
  • Added control for more precise object geometry as detected by general occupancy network.
  • Improved control for vehicles cutting out of our desired path by better modeling of their turning / lateral maneuvers thus avoiding unnatural slowdowns.
  • Improved longitudinal control while offsetting around static obstacles by searching over feasible vehicle motion profiles
  • Improved longitudinal control smoothness for in-lane vehicles during high relative velocity scenarios by also considering relative acceleration in the trajectory optimization.
  • Reduced best case object photon-to-control system latency by 26% through adaptive planner scheduling, restructuring of trajectory selection, and parallelizing perception compute. This allows us to make quicker decisions and improves reaction time.

© 2022, Eva Fox | Tesmanian. All rights reserved.

_____________________________

We appreciate your readership! Please share your thoughts in the comment section below. 

Article edited by @SmokeyShorts; follow him on Twitter








← Vorheriger  / Nächster →