Tesla Rolls Out FSD Beta 10.69.3 on Halloween Night

Tesla rolled out Full Self-Driving (FSD) Beta 10.69.3 with release notes available online on Halloween night showing a massive list of major changes.

Several weeks before the FSD Beta release, Elon Musk tweeted the coming of FSD Beta. It came out, indeed, as promised and began to appear on Halloween night.

FSD Beta 10.69.3, according to Not A Tesla App, rolled out with 2022.36.15. That means the beta includes the features of Tesla's 2022.36 update alongside Cabin Overheat, the new Energy app, and hosts of other exciting features.

Driver-Assist System

The FSD Beta empowers the vehicle to drive independently to a destination encoded in its navigation app. The driver in the driver's seat, however, has to be vigilant and ready to take over the vehicle if the need arises.

Basically, it is still a driver-assist system with the driver taking on full responsibility and not the Tesla system.

Despite the name, according to Electrek, the vehicle is a type of "two steps forward, one step back" program.

One or Two More Updates.

More owners have been added to the FSD Beta program as the carmaker known for its electric vehicles has the practice of releasing software updates.

More than 100,000 people are now in the Beta program. Electrek said the company plans to add more owners by year-end with more updates to fine-tune the system.

Normally, Tesla delivers a new Beta update once every month. The company will likely come up with one or two more updates from the expected wider release.

Beta Testers Next

The current FSD Beta 10.69.3 is intended for internal testing among Tesla employees. Logically, the company will release Beta testers soon to expand the testing to customers.

Based on the release notes, Electrek observed the software update did not include new features but offered many high-level updates to boost overall performance.

Below is the massive list of features of Tesla Full Self-Driving Beta Release Notes v10.69.3 Release Notes via Not a Tesla App:

- Upgraded the Object Detection network to photon count video streams and retrained all parameters with the latest auto-labeled datasets. A special emphasis was 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 the 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 retrained 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 hillcrest.

- Reduced runtime of candidate trajectory generation by approximately 80% and improved smoothness by distilling an expensive trajectory optimization procedure into a lightweight planner neural network.

- Improved decision-making for short-deadline lane changes around gores by richer modeling of the trade-off between going off-route versus the trajectory required to drive through the gore region.

- Reduced false slowdowns for pedestrians in crosswalks by using a better model for the kinematics of the pedestrian.

- Added control for more precise object geometry as detected by the 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 considering relative acceleration in the trajectory optimization.

- Reduced best-case object photon-to-control system latency by 26% via adaptive planner scheduling, restructuring of trajectory selection, and parallelizing perception computation. This feature allows quicker decisions and improves reaction time.

The release of FSD Beta 10.69.3 will give Tesla more listening time to beta drivers about possible concerns they may raise.

Tesla, according to a report by Teslarati, is working on the wide release of FSD Beta with the last quarter.