Documents rus
Main page / Blog / To autopilot or not to autopilot

To autopilot or not to autopilot

23.01.2018
Andrey Nikishin

On January 22, an article was published that described an unusual car accident at 65 miles per hour, which, fortunately, did not lead to any injuries. The accident occurred when a Tesla Model S with on-board autopilot enabled crashed into a fire truck, which at the time was stationary on the freeway assisting the victims of another accident.

An investigation is currently underway and it’s not completely clear whether this is an autopilot error or the driver is trying to push the blame onto the car. This is definitely not the first accident involving autopiloted Teslas, which is why the technology cannot be described as completely reliable. But if you think how many miles cars have done using autopilot without crashing, then your view may change.

It seems to me that the drivers of Tesla are now so confident in autopilot’s ability that they probably get distracted and are not always paying attention to the road. Why would a driver keep his hands on the wheel and watch the road carefully if autopilot has not let anyone down and cars with autopilot enabled had traveled more than 130 million miles? Before the accidents started happening, everything was going well. Maybe we overestimate the development of autonomous cars and put too much trust in immature technology. In my opinion, it’s too early to call the current generation of autopilot controls the finished product. It’s more of a super adaptive cruise control, or driver assistant. It’s a system, but not autopilot. After all, the word autopilot means the car can travel on its own, automatically and without any kind of interaction from a person. Cruise control regulates the speed, distance, and lane, but everyone knows you have to keep your hands on the steering wheel and watch where you’re going.

Modern cars are getting smarter and smarter, thereby increasing safety on the roads and autopilot is one of the technologies that will radically affect the quality of fully autonomous transport – something that should happen quite soon. Automakers promise that by 2020 there will be a lot of cars using autopilot on the roads. I would like to hope that in 3-5 years, the technology will reach such a level that we won’t be afraid of it, but for now, I believe we shouldn’t overestimate it.

We know that the rules of road and aviation safety are written in blood. I would like to hope that the bloody page of rules devoted to autopilots will be as short as possible.