Analysis: Self-driving tech is beset with difficulties

Car makers are continually advancing their driver assistance tech, but issues remain.

By Jim Holder, Autocar UK calendar 10 Nov 2020 Views icon3325 Views Share - Share to Facebook Share to Twitter Share to LinkedIn Share to Whatsapp
Analysis: Self-driving tech is beset with difficulties

Perhaps the most complex challenge facing the car industry today is vehicle autonomy, the path to achieving it and the motivations for doing so.

To most, the end goal isn’t in question; a utopia of zero accidents is too compelling for legislators to ignore and the rewards of success (this is an industry estimated to be worth £5.5 trillion (Rs 40,705,500 crore by 2050 by the admittedly invested Intel) too great to be overlooked by established firms or disruptors.

While some of the tech is there or thereabouts now, that in itself is part of the problem: unless it delivers perfection, autonomous tech is fallible, just like human beings, and thereby open to all sorts of moral, emotional and legal challenges.

Supporters will point out that today’s optimistically labelled ‘self-driving’ tech has already proved its value, with accidents occurring far less than when people drive unaided. This is a line that Tesla boss Elon Musk has espoused before, although his data was challenged.

Others will suggest that the cost in terms of injury and death is worth it in the short term in order to advance the tech faster for the long-term. It’s a logical argument but no comfort to the parents of a child killed by a computer’s failure.

To degrees, many new-car drivers already experience autonomous tech’s abilities when they use adaptive cruise control or lane-keeping assistance. While these often make for a jerky and unsettled drive, they can on a good day get you from London to Wales on the M4 with scarcely an input.

Tech that works in ideal conditions is all well and good, but what happens when it’s impaired, say by the weather or dirt? Or when something unpredictable happens ahead and the car is moving at 70mph / 110kph? Can a computer really read the scenario as a human can?

And even if it can, how should it react? Will brands with safety at their heart programme their tech to be warier than those that prioritise driving thrills? For instance, would a Volvo pull up at the first sign of trouble but a BMW wait a moment longer? If they all offer the same thing, why choose one over another?

And then there are the issues of who is at fault if it does go wrong. The tech developer who created the systems, the car maker who sold the tech or the driver who elected to switch it off or on? Or do insurers have to carry the can, even during the transitional stage when some cars might run the systems and others won’t?

As the saying goes, if it were easy, everyone would be doing it. But as with flying cars, there’s always the risk that ambition and reality don’t mix.

READ MORE

Self-driving Ford Mondeos roll out in Oxford for government-backed project 

Survey finds majority of UK drivers not ready for driverless cars 

New ‘deepfake’ tech to accelerate autonomous car development

 

RELATED ARTICLES
Volkswagen Group sells 465,500 BEVs worldwide in first-half 2025, up 47%

auther Autocar Professional Bureau calendar10 Jul 2025

With strong growth in Europe (+89%) and the USA (+24%), despite a sales decline in China (-34%), the VW Group’s global B...

Skoda begins sale of made-in-India CKD Kushaq in Vietnam

auther Autocar Professional Bureau calendar30 Jun 2025

Before production started, pre-series Kushaq vehicles covered over 330,000 kilometres on a variety of Vietnamese roads a...

Six Japanese companies join forces to expand use of recycled materials in new vehicles

auther Autocar Professional Bureau calendar30 Jun 2025

Denso, Toray Industries, Nomura Research Institute, Honda Motor, Matec Inc and Rever Corporation have set up the BlueReb...