As we learned from a new report by the National safety Board of transportation, the sedan Tesla Model 3, which in March crashed into a truck on a highway in Florida, and killed his driver, had acted with the active function of the autopilot. This driver became at least the fourth died in an accident related to the function of semi-Autonomous driving Tesla. What is striking in the crash on March 1, is the fact that the details of the accident are almost identical to the details of the first public reports of a fatal crash the autopilot in may 2016.
In each case, Tesla on autopilot collides with a truck, crossing the road, killing the driver.
Victims of autopilot Tesla Model 3
CEO Elon Musk boasted powers of self-government of their cars and promised that next year they will become fully Autonomous. But critics say the autopilot system — which requires that the drivers were alert and ready to take over the management of the car — works well enough to keep drivers in check, and that Tesla produces a system more capable than it really is. New crash casts doubt on how well Tesla has responded to this criticism, given its active promotion of more ambitious technologies.
In the March crash of a red Tesla Model 3 was traveling South in the right lane of state road 441 in Delray beach, about 110 miles North of Miami. The truck left from a private road on the right side of the highway, intending to turn left and head North. The report says that the truck slowed down, crossing the lane, “blocking the way Tesla”.
Autopilot system for Model 3 included about 10 seconds before the accident and the car didn’t see the driver’s hands on the steering wheel, within eight seconds, immediately preceding impact. The representative of the Tesla confirms this. The car crashed into a trailer at a speed of 110 km/h (if speed limit is 90 km/h), not trying to make any manoeuvres of evasion. The car drove under the truck, tearing off the roof and killing the driver Jeremy Beren Banner, age 50 years. The truck driver was not injured.
In the report it is not reported how far was the Tesla when the truck left the road, and not marked by weather conditions that could affect the vehicle’s ability to brake. It is therefore unclear could the driver not using the autopilot, to stay safe. Rough math suggests that the driver Model 3 would require a few seconds to assess the situation and avoid an accident. At a given speed the car travels 30 meters per second. The braking distance Model 3 at a speed of 100 km/h is about 40 feet. The car could stop in 75 meters. Add 1.5 seconds to the actions of the driver by braking and, like, three seconds would be enough.
May 7, 2016 40-year-old Josh brown died under very similar circumstances. Its Model S was working on autopilot as he drove North on highway 27A in Northern Florida. Towards drove the truck. Tesla stopped traffic and drove at the speed of 125 km/h (with speed limit of 110 km/h), held under the truck, tore the roof and killed brown. Then the car traveled another 100 meters, hard enough hit the pole and through 15 meters stopped. The truck driver was not injured.
Then it became obvious that the truck driver had to give Tesla, and brown was inattentive. But some of the blame rests on Tesla, which allowed the driver to over-rely on the automation system over a long period.
After the death of brown, Tesla has reduced the time during which the driver can not touch the steering wheel before the system will give audible and visual warnings. Went through a few iterations of hardware and software. But the circumstances of the accident indicate that when driving at high speed the system is still not capable of detecting some stationary objects or objects moving perpendicular to the car. Probably for the same reason, at least three Tesla crashed into standing fire trucks in 2018 (without serious injury).
Model 3, which came in the last accident, was equipped with cameras that were, theoretically, to detect the truck crossing the path of the machine. Tesla did not respond to a request for Wired about how the system uses radar and camera or what steps have been taken to avoid such accidents.
At the same time, Tesla optimistic about the capabilities of machine vision camera technology, which should allow “full self-driving”, according to the Mask, without the need for human observation. Last month, head of technical Department at Tesla Autopulot Andrew Karpati stated that Tesla believes that it can use machine learning methods and camera to take their cars for quality drivers. Musk went further and said: “I quite confidently predict the emergence of an Autonomous, robotaxi from Tesla in the next year.”