As Tesla gears up for its earnings release on July 23, all eyes are on the company’s ambitious Robotaxi platform expansion. Elon Musk has promised a significant improvement in the upcoming Full Self-Driving (FSD) release, drawing from valuable insights gained from the driverless ride-share service.
In Austin, a select fleet of Model Y vehicles has been converted into robotaxis, operating with a unique FSD software distinct from the version used by everyday Tesla owners. Musk has revealed plans to merge this specialized FSD branch with the commercial software, exercising great caution in the process.
During its initial launch in Austin, Tesla described the Robotaxi fleet’s FSD as “unsupervised,” in contrast to its paid driver assist option, which requires supervision. The Austin transportation department has classified it as a Level 4 autonomous vehicle, capable of functioning independently of a human driver.
To gain regulatory approval, Tesla employs a different operational approach for its Model Y robotaxis compared to the public FSD branch. For instance, the Robotaxi platform operates within a geofenced area, similar to Waymo’s strategy, ensuring these “unsupervised” vehicles navigate routes consistently with other Tesla models.
This approach explains the numerous Tesla validation vehicles seen around Austin, equipped with additional camera setups, before the Robotaxi service launch. Before integrating the robotaxi FSD with the general public version, Tesla plans further testing to guarantee a seamless upgrade. “We need to validate that improvements for Austin don’t cause regressions elsewhere,” Musk clarified.
Tesla is renowned for its camera-only FSD strategy, emphasizing spatial freedom and scalability over the extra sensor and LiDAR solutions used by competitors like Waymo. However, for safety and regulatory reasons, Tesla still employs geofencing in Austin, ensuring its robotaxis behave similarly to Tesla vehicles on public roads outside predetermined zones.
The integration of the robotaxi FSD with the general release is anticipated later this year. Musk has hinted at a fourfold increase in parameters and a highly efficient use of memory bandwidth, optimizing computation for the additional algorithms.
