Could 2021 be the year of the driverless car? Just a few weeks ago, the snappily named Xpeng P7 Super Long Range Premium hit the road with autonomous tech that was billed as some of the most advanced ever to grace a production car. And, of course, this is the year that Ford planned to have its driverless taxi fleet up and running, while Google’s automotive offshoot, Waymo, predicted it’d have around 100,000 self-driving cars on the road by now. And yet…

The truth is we seem to be further away from a driverless reality than ever. Now, obviously, the Covid pandemic has hobbled all areas of the industry, with Ford claiming this is the reason its automated ride-hailing service has been postponed. Confidence has been further eroded by Uber’s sale of its self-driving division earlier this year for half its value. And Waymo? Well, it does have an operational autonomous fleet, but it currently numbers just 600 vehicles or so.

However, could there be a fundamental flaw at the heart of autonomy that has nothing to do with technology, investment or pandemics? Could the biggest hurdle to success actually be the very thing that driverless cars are trying to eliminate: you and me?

How so? Well, for most observers and experts, the accepted view of the medium-term future of this technology is that self-driving machines should still be capable of manual override, such as when the road ahead gets interesting and you want to get involved or, less likely maybe, when the technology fails and requires help.

It’s this last scenario that could prove the most challenging because, let’s face it, the better the tech gets, the less we’ll drive, which means when the car does have a high-tech hiccup, we’re likely to be somewhat unprepared. As far back as 1983, esteemed psychologist Lisanne Bainbridge had written a study that shone a light on the potential pitfalls of relying on heavily automated systems. In short, she concluded that the more we rely on the tech, the worse we become at the task it’s assisting us with.

In the case of cars, most industry insiders reckon a worst-case scenario for a failure that requires intervention from the driver would be classed as a once-in-every-200,000-miles-or-so event. Sounds impressive, doesn’t it? I mean, if you were told you wouldn’t experience any sort of breakdown for at least 200,000 miles, you would be overjoyed. Until you realise that you could have changed cars three or four times in that period and probably won't have actually ‘driven’ for maybe five years. And now the black box wants you to take over, more than likely in a car whose controls you’re totally unfamiliar with. That’s a fairly sobering thought.

It’s not just that you’ll be out of practice, because there’s also the small matter of reaction times. Simulations have shown that once the driver is occupied with something else, it takes around 40 seconds from the moment the car requests they take over to the point at which they’ve fully got a handle on their surroundings and what’s required. Imagine how far you’d have travelled in that time at 70mph.

So for the technology to be truly ‘safe’, you’d have to be alert and engaged the whole time, which sort of defeats the object of the exercise. With time, effort and money, even this problem will be overcome, but it’s a way off yet. So, no, 2021 probably won’t be the year of the driverless car.