started · updated
Tesla Full Self-Driving software faces safety challenges in recent tests
Recent tests and incidents involving Tesla’s Full Self-Driving (FSD) software have highlighted both its capabilities and significant safety limitations. In a documented incident, a Tesla running the latest FSD v14.3.7 software nearly collided with a train after failing to stop at an active railroad crossing despite flashing signals and audible horns. The driver had to intervene manually to avoid the tracks.
In separate testing environments, the technology faced different challenges. A Tesla Model 3 was used to navigate Lick Skillet Road in Colorado, an unpaved mountain route with steep grades of up to 18 percent and no lane markings. Additionally, a week-long evaluation of a Model Y revealed that while the system can manage highways and urban streets, it still struggles with lane selection, speed limit recognition, and managing right-of-way in complex intersections.
Despite Tesla’s classification of the system as supervised driving, these instances underscore the ongoing necessity for constant driver vigilance to prevent accidents caused by software errors in varied terrain or at critical infrastructure crossings.
Entities
Claims
What the coverage asserts, and how many sources carry each claim.
- [○ 1 SOURCE] Lick Skillet Road has an average slope of 14.2 percent with sections reaching 18 percent. www.autojournal.fr
- [○ 1 SOURCE] A Tesla Model 3 using Full Self-Driving (Supervised) navigated Lick Skillet Road, a steep dirt road in Colorado. www.autojournal.fr
- [○ 1 SOURCE] A Tesla Model Y test revealed errors in lane selection, speed limit reading, and managing right-of-way. www.altroconsumo.it
- [○ 1 SOURCE] The Tesla was approaching the crossing at 9 mph with the Self-Driving indicator active when the incident occurred. electrek.co
- [○ 1 SOURCE] A Tesla running FSD v14.3.7 failed to stop at an active railroad crossing and nearly drove into a passing train. electrek.co