Tesla’s latest Full Self Driving software has just started rolling out to customer cars in the US.
The build version is Version 2023.7.20, which attracts the FSD Beta build number of V11.4.4.
At the time of writing, the build is installed on 54 vehicles on TeslaFi and 42 on TeslaScope. With every new build of the software, we expect progress to resolve issues and improve the human-like behaviour of driving.
What’s interesting about the release to customer vehicles today, is that it was just yesterday when we first learned this build had reached employee vehicles. Historically the latest releases are deployed to employee cars for at least for a few days, but up to a week before receiving a wider rollout.
One possible explanation for this accelerated release is that changes are minor and Tesla completed testing quickly and feels comfortable deploying. Another possible explanation is that Tesla saw a bug in a previous build that was important to update to resolve.
As we know, Australia is awaiting the release of FSD Beta. Things are moving quickly in this space, with as many as 13 vehicles in Australia now reporting as having a software build (2023.3.6) with FSD Beta on TeslaScope. Currently, we believe these are still Tesla employee and company cars and no customers in Australia have received a build, but we wait and hope this isn’t far away.
FSD Beta 11.4.4 Release Notes
While we don’t yet have the full release notes, we do have a portion of them, which are available below. These show some improvements that hopefully resolve some issues in the software.
– Improved short-deadline lane changes, to avoid going off-route, through better modeling of target lane vehicles to improve gap selection assertiveness.
– Improved offset consistency when controlling for static obstacles. Also improved smoothness when changing offset direction by adjusting speed more comfortably.
– Improved handling of oncoming cars on narrow unmarked roads by improving prediction of oncoming car’s trajectory and leaving enough room for them to pass before re-centering.
– Improved Occupancy Flow prediction from the Occupancy Network for arbitrary moving obstacles by 8%.
– Expanded usage of the new object ground truth autolabeler for the NonVRU detection model, improving distant vehicle recall and geometry precision for semi-trucks, trailers, and exotic vehicles.
– Improved VRU control by expanding planning scope to control gently for low-confidence detections that may interfere with ego’s path.
– Improved handling for VRUs near crosswalks by predicting their future intent more accurately. This was done by leveraging more kinematic data to improve association between crosswalks and VRUs.
– Improved ego’s behavior near VRUs by tuning their assumed kinematic properties and utilizing available semantic information to classify more accurately their probability of intersecting ego’s path.
– Improved Automatic Emergency Braking recall in response to cut-in vehicles and vehicles behind ego while reversing.
