The idea of a self driving car has been in the minds of every Sci-Fi fan for years. We all wait for the day when we can take a long journey just sitting back without worrying about driving. But, why are we so obsessed with autonomous vehicles? For transport corporations the answer is obvious. If there is no driver there is no salary to pay and that way operational costs will be lower.
But what about us, the ones not worrying about cutting cost to increase profits. Why do we embrace and can not wait until self driving cars are a reality. First of all human beings are usually bad drivers, we get distracted, we drink, use our phones or even fall asleep in long journeys. Basically our human condition is preventing us from becoming the perfect drivers.
If you can not beat them…Let them take over.
An AI (Artificial Intelligence) is a better candidate to drive us around without making the same mistakes that we do while driving. Around 94% of the US traffic accidents are caused by human error. But is it possible for an AI to take over completely and make the hard choices when the time comes? Most car manufacturers seem to think so, at least in the near future.
These cars are equipped with sensors designed to detect objects as far as two football fields away in all directions, including pedestrians, cyclists and vehicles. All the info gathered is then processed by the AI software to help the car safely navigate the road without getting tired or distracted.
According to TESLA, all the vehicles produced in their factory will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver.
Sounds good… I want one.
Well, not so fast. Although you can buy a TESLA equipped with all the hardware, the self driving capability is still not enabled. The system will need further calibration and millions of miles of real world driving in order to ensure significant improvements to safety and convenience. For now the closest we get to the self driving experience is the lane holding and active cruise control with emergency braking and collision warning.
So… No Competition?
Although most, if not all, car manufacturers (GM, Daimler, Volvo, Ford, Jaguar Land Rover, Audi, BMW) are jumping on the self driving car wagon, they are yet to release such a consumer vehicle. Google and Uber have their own experiences with self driving cars, but are mainly focusing in the ride sharing business.
Apple to tried to get into this space too, of course, but has recently announced laying off dozen of employees and even closing down part of the project. Instead, it’s said to focus on developing autonomous navigation software, rather than a vehicle.
Recently a new start up (comma.ai) was promising to sell an off the shelf self driving module. Any user with a compatible consumer vehicle could buy the module and install it himself. It was supposed to start shipping later this year, but legal implications are preventing them from selling the device for the US market.
Mobileeye was working on a self driving car but is now focusing on developing collision avoidance features for consumer vehicles.
Udacity, an e-learning company, announced recently an interesting project. It has the objective of building an open source self driving software with the help of the community. It still is in its early stages but has already gathered a lot of support.
Why can’t we have it now?
The technology exists and has been rigorously tested for a while. But it’s still not at a stage where we can rely on the software to drive us around. There are legal implications that will need to be settled and insurance companies are paying close attention to this field.
Regulations need to be thoroughly thought through, discussed, introduced and most importantly addressed before the cars can hit the roads. Just two months ago the US government released 114-page document governing the manufacture and sale of self driving cars. Tesla autopilot crash that happened earlier this year is still being investigated, but no doubt many more regulations will need to be introduced before self driving cars get into mass market.
Some ethical implications on the decision taking of an AI driving a potential deadly machine are also concerning. What decision would the AI take in case of eminent danger? Would it protect its occupants or pedestrians? Would it prioritise children over elderly? Mit developed an interesting project called Moral Machine. It identifies various situations a self-driving car might find itself in and you can ‘experience’ an ethical dilemma your future car will.
Self driving cars are yet at a point of fascination, consumer excitement, manufacturer research and competition amongst brands. Once self driving cars get seriously close to hitting the roads (and yes, we’ve already come across Uber’s self driving car around Hoxton, which, by the way, nearly hit a cyclist) – a serious public debate on its moral implications will be needed. Who should AI protect? What about human psychology – can it be changed as quickly? Only after we understand the implications better and accept the moral risk, will commute to work on the back seat of a self driving car become a reality.