Stories about autonomous vehicles, self-driving cars and trucks, have been splashed all over the news and social media, with major companies such as Tesla, Uber and Google trying to be the first to corner the market. But for all the hype, there are many reasons to believe it will be quite a while before autonomous vehicles become the dominant kind of transport seen on our roads.
They could be hacked
The fact that autonomous vehicles are connected via wireless internet, to allow them to pass crucial information back and forth, leaves them vulnerable to hackers.
It is feasible (and in fact, has been quite clearly demonstrated) that a hacker could remotely interfere with the function of vital systems inside an autonomous vehicle, and could even disable the brakes. Someone could murder you via your self-driving car, and make it look like a computational malfunction.
Until highly robust cyber security systems have been developed to ensure a very low risk of autonomous vehicles being hacked, the chances of abuse will be too high for most to endure.
The public doesn’t trust them
Research on driverless vehicle opinions by Intelligent Car Leasing, looking at public attitudes in the UK, found that 61% of respondents would feel safer in a vehicle driven by a human than one controlled by an AI system. Of the 1750 people aged between 18 and 65 who were surveyed, only 17% said they would feel safer in the self-driving vehicle, with 22% indifferent.
Until the autonomous vehicle industry is able to convince a majority of the public that its products are as safe to use as normal human driven vehicles, autonomous vehicles will not take over. It is possible that a gradual growth in public acceptance could lead to a tipping point in which a majority favourable to the vehicles rapidly appears, but it’s unknown how long it would take to reach such a tipping point.
They still can’t handle tricky situations
A few recent incidents show that autonomous cars are still not able to handle all difficult situations where the lives of human beings are at stake.
In early 2018, a woman in Arizona, USA became the first pedestrian to be killed by an autonomous vehicle. She was crossing the road when she was fatally struck at high speed by an Uber self-driving car. What is especially concerning is that the car’s computer systems spotted her, but for some reason did not initiate an emergency stop.
In the same year, the driver of a semi-autonomous Tesla car running in “autopilot mode” was killed when the car failed to spot a truck passing horizontally in front of it. It is believed that the computer vision systems in the Tesla were confused by the sun glinting off the metal side of the truck.
While incidents such as these are rare, they demonstrate that the collision avoidance systems of autonomous vehicles have still not reached a high enough standard.