FYI: Tesla’s Full-Self Driving (Supervised) system shows promise but struggles with reliability over extensive testing.
Tesla’s Full-Self Driving (Supervised) System Put to the Test: 1,000 Miles with Mixed Results
Tesla’s much-heralded Full-Self Driving (Supervised) advanced driver assistance system (ADAS) underwent rigorous testing, and the results were a mixed bag. Conducted by AMCI, an independent automotive research firm, the comprehensive review spanned over 1,000 miles, revealing notable insights and areas of concern. Despite Tesla’s bold assertions about self-driving capabilities, the technology still hasn’t reached the level of autonomy that many enthusiasts and consumers are hoping for.
### The Testing Process and Results
AMCI Testing put the Full-Self Driving (Supervised) system through its paces on diverse terrains, including city streets, rural highways, mountain roads, and major highways. The 2024 Tesla Model 3 Performance used in the tests was equipped with the latest hardware and software versions 12.5.1 and 12.5.3. Despite this, drivers had to intervene over 75 times, resulting in an average of one intervention every 13 miles.
Instances requiring driver intervention included the car running a red light at night, veering into oncoming traffic when navigating a twisty rural road, and stopping inexplicably at a green traffic light. These incidents underscore the unpredictable nature of the system, which can perform flawlessly in similar scenarios multiple times before encountering a glitch.
### Expert Insights
Guy Mangiamele, Director of AMCI Testing, commented on the unpredictability of the system: “What’s most disconcerting and unpredictable is that you may watch FSD successfully negotiate a specific scenario many times–often on the same stretch of road or intersection–only to have it inexplicably fail the next time.”
AMCI also released several videos documenting where the system performed well, such as moving aside on narrow roads to let oncoming cars pass, and where it failed. These clips highlight both the potential and the pitfalls of Tesla’s technology.
### The Double-Edged Sword of Near-Perfect Performance
David Stokols, CEO of AMCI Global, weighed in on the implications of near-perfect performance: “With all hands-free augmented driving systems, and even more so with driverless autonomous vehicles, there is a compact of trust between the technology and the public. Getting close to foolproof, yet falling short, creates an insidious and unsafe operator complacency issue as proven in the test results.”
### The Road Ahead
These findings are particularly timely as Tesla prepares to launch its Robotaxi service on October 10. Despite claims by CEO Elon Musk that these cabs will be able to drive autonomously without pre-mapped data, real-world testing suggests there may still be hurdles to overcome. Bloomberg and Tesla hacker Green The Only have reported that Tesla is actively collecting data in Los Angeles, where the launch event is set to take place. Observant Reddit users have even spotted test vehicles on the roads where a bright yellow mule resembling a two-door Cybercab was photographed.
### Conclusion
While Tesla’s Full-Self Driving (Supervised) system shows significant promise, the testing by AMCI indicates that it’s not yet ready to eliminate the need for human oversight. The occasional but critical failures highlighted by the research underscore the ongoing challenges in achieving truly autonomous driving.
For more insights and updates, reach out to us at tips@automotive.fyi, or follow us on Twitter @automotivefyi.
William Kouch,
Editor of Automotive.fyi
