Tesla Raises Concerns of Putting Cost ahead of Safety
Tesla has reignited the debate over driverless technology requirements by deciding to focus on camera-only Level 2 automated driving systems.
Reuters has reported that the BEV automaker has now dropped radar sensors from its semi-autonomous driving system, Autopilot, raising concerns over the safety of the camera-only version, Tesla Vision. On the face of it, the system fitted to new Model 3 and Model Y cars will fail to meet existing safety standards both in the US and Europe.
Already, the National Highway Traffic Safety Administration (NHTSA) has withdrew its advanced safety features label for these vehicles and Consumer Reports dropped its “top pick” label. The cars are also likely to lose their NCAP safety ratings until the camera-only system has been fully tested.
The move has widely raised concerns particularly following increasing instances of crashes implicating possible failures of Autopilot resulting in death and serious injury. Latest in this was the death of two men found in a crashed Model S with neither of them sat in the driver’s seat illustrating how the current system can be fooled.
Cost saving
Naturally, many experts believe Tesla’s move is more about cost savings involved by using more expensive sensors such as radar and LiDAR over the comparatively cheaper option of cameras reliant on improved software.
Among experts quoted by Reuters, Raj Rajkumar, professor of electrical and computer engineering at Carnegie Mellon University, said: “You need to use all the different kinds of sensors and then combine them.” Steven Shladover, a research engineer at Berkeley University of California, said losing radar in ADAS is enough “to render them less usable to unusable in adverse weather conditions”. He added: “It makes no sense whatsoever technologically – only a way of reducing cost of components.”
Ram Machness, chief business officer of advanced radar maker Arbe Robotics, warned the loss of radar could adversely affect established safety features such as emergency braking to avoid accident situations.
— Paul Myles is a seasoned automotive journalist based in Europe. Follow him on Twitter @Paulmyles_
Under the point of technology I do believe autonomous conduction would be a reality within some years from now.
Although I’ve serious reserves about his acceptation after revealed the price difference for the option.
Indeed the minor of people who can afford to have a very expensive car pay to have a personal driver, in the middle we have people who like to drive, to feel the car.
For those people a power and fast car it’s a passion they love to drive to prove himself they are a good driver, they don’t like anybody-else to drive his car.
Then we have those people who have a car as a “heavy” investment to bring the family out for a lunch or to the beach, for them any extra screw it’s not justified for a car used for weekends.
And for today’s technology “autonomous” conduction to be save in any circumstance represents significative costs both in terms of development, production and heavy fees for guaranty and insurance
Nobody can hope to have an AV for a price just a bit higher than the average, to buy and to maintain.