UPDATE: Tesla’s Full Self-Driving (FSD) software faced a significant challenge navigating a multi-story carpark at Albury’s Commercial Club, home to one of Australia’s largest Supercharger stations. This test, conducted today, aimed to evaluate FSD’s capability in complex parking environments.
The Albury Supercharger is located on the first floor, requiring vehicles to navigate a series of turns and a steep ramp to reach the charging stations. This capability is critical not only for Tesla owners but also for electric vehicle (EV) users who frequently park in multi-story facilities, such as shopping centers and hotels.
Upon entering the destination into the navigation system, the Tesla initially parked on the street instead of entering the carpark. This raised immediate concerns about its navigation capabilities. After manually driving into the carpark, the tester engaged FSD, which surprisingly attempted to exit rather than complete the intended task.
Re-entering the carpark, the tester reactivated FSD as they approached the on-ramp. The vehicle managed to ascend the ramp but came to a halt as it reached a critical point where visibility was limited. The FSD indicator shrank, indicating a lack of confidence in proceeding. A quick tap of the accelerator allowed the car to gain sufficient visibility and reach the Supercharger level.
However, instead of parking at the Supercharger, the car continued to navigate to the top floor of the carpark. Once it reached the coordinates of the Supercharger, it entered free-roam mode, failing to complete the parking task. This highlights a significant limitation in FSD’s current navigation capabilities, particularly when correctly interpreting vertical navigation in multi-story structures.
After charging, the tester set the destination for the exit. In a notable turnaround, the FSD successfully navigated the exit ramp and made the necessary turns to leave the carpark. Yet, during this exit, the vehicle incorrectly chose a left turn instead of a right, which could have led to a more significant issue if traffic were present.
The tester noted that the car’s ability to detect obstacles like pedestrians and vehicles was satisfactory, but the software needs to improve its navigation confidence, particularly in low-visibility situations. Furthermore, the vehicle’s navigation system appeared to rely solely on two-dimensional coordinates without fully accounting for the vertical aspects of parking structures.
This evaluation underscores the ongoing challenges Tesla faces in refining its FSD technology, particularly for complex environments. As the demand for EVs continues to rise, the ability to navigate multi-story carparks will be crucial for Tesla’s market competitiveness.
With these findings, Tesla may need to enhance its navigation algorithms to better handle vertical navigation and improve user confidence in its FSD system. As developments unfold, the implications for Tesla and the broader EV market remain significant.
For a firsthand look at the experience, be sure to check out the accompanying video.
