Based on recent comments from Elon Musk, Full Self-Driving (FSD) Beta testers were expecting the next release to be version 10.12. It was also expected that the new version would be released earlier this week.
As we already know, the release was delayed to “this weekend”, and it looks like it has met that timeline, but only internally within Tesla.
According to Tesla enthusiast @WholeMarsBlog, 10.11 is currently being tested by employees, and if no significant bugs are found, could be released to external users on Monday.
FSD Beta 10.11 has gone out to Tesla employees.
If all goes well, external users should get it tomorrow. @elonmusk
— Whole Mars Catalog (@WholeMarsBlog) March 13, 2022
Shortly after revealing this information, @WholeMarsBlog was provided with the 10.11 release notes, which indicate a number of substantial improvements. Tesla has been quantifying the rate of improvements in their FSD Beta release notes lately, but we have never seen improvements in the magnitude of 63.6%.
The overall theme of this release appears to be at reducing unnecessary slowdowns.
ok got them lol pic.twitter.com/NXfjNPox8J
— Whole Mars Catalog (@WholeMarsBlog) March 13, 2022
The big question that remains is whether Canada will get this next release at the same time as our friends south of the border.
The initial timeline provided by Musk was that there would be a 3 day gap in releases between both countries, but it is unknown if the delay has cancelled the need for that gap.
Tesla owners in Canada are hoping there will be no gap so that they no longer have to nurse their 100 Safety Score.
FSD Beta 10.11 Release Notes (2022.4.5.15)
Upgraded modeling of lane geometry from dense rasters (“bag of points”) to an autoregressive decoder that directly predicts and connects “vector space” lanes point by point using a transformer neural network. This enables us to predict crossing lanes, allows computationally cheaper and less error prone post-processing, and paves the way for predicting many other signals and their relationships jointly and end-to-end.
Use more accurate predictions of where vehicles are turning or merging to reduce unnecessary slowdowns for vehicles that will not cross our path.
Improved right-of-way understanding if the map is inaccurate or the car cannot follow the navigation. In particular, modeling intersection extents is now entirely based on network predictions and no longer uses map-based heuristics.
Improved the precision of VRU detections by 44.9%, dramatically reducing spurious false positive pedestrians and bicycles (especially around tar seams, skid marks, and rain drops). This was accomplished by increasing the data size of the next-gen autolabeler, training network parameters that were previously frozen, and modifying the network loss functions. We find that this decreases the incidence of VRU-related false slowdowns.
Reduced the predicted velocity error of very close-by motorcycles, scooters, wheelchairs, and pedestrians by 63.6%. To do this, we introduced a new dataset of simulated adversarial high speed VI interactions. This update improves autopilot control around fast-moving and cutting-in VRUs.
Improved creeping profile with higher jerk when creeping stand ends.
Improved control for nearby obstacles by predicting continuous distance to static geometry with the general static obstacle network.
Reduced vehicle “parked” attribute error rate by 17%, achieve increasing the dataset size by 14%. Also improved brake light accuracy.
Improved clear-to-go scenario velocity error by 5% and highway scenario velocity error by 10%, achieved by tuning loss functions targeted at improving performance in difficult scenarios.
Improved detection and control for open car doors.
Improved smoothness through turns by using an optimization based approach to decide which road lines are irrelevant for c ? given lateral and longitudinal acceleration and jerk limits as w? vehicle kinematics.
Improved stability of the FSD Ul visualizations by optimizing ethernet data transfer pipeline by 15%.
Improved recall for vehicles directly behind ego, and improved precision for vehicle detection network.
