AI Online


In December 2015 California motor vehicle department officials proposed self-driving regulations that make it compulsory for a real live driver to be able to take control where necessary – in other words, the car has a steering wheel, brakes and accelerator.

As perhaps something of a traditionalist this writer sees nothing unusual in the proposal. However, Google in response to an AFP enquiry stated: “We’re gravely disappointed that California is already writing a ceiling on the potential for fully self-driving cars to help all of us who live here”. It can be argued that California set the trend towards the lowering of emissions in 1966 when tailpipe emission standards for hydrocarbon and carbon monoxide emissions were adopted by the California Motor Vehicle Pollution Control Board. These were the first of their kind in the United States, and one of the first in the world. It is, therefore, safe to assume that other states and countries will be taking their cue from the Californian rules of the road for self-driving cars.

It may be a good idea for both the legislators and the technical teams designing self-driving or autonomous vehicles for the roads to first ask the customer what they want. One of the predictions for the coming year by consulting and technology firm Capgemini is that 2016 will see the first driverless vehicle accident caused by a software fault.

In this edition of Automotive Industries we have a number of companies discussing another threat – that of cyber security. A control-less car or one without a mechanical override will be at the mercy of hackers. Another challenge which does not seem to be gaining much traction as yet is the reliability of wireless connections. It is not unheard of that cellular and other networks go down. Autonomous or semi-autonomous vehicles rely on constant streams of data on which to base calculations. Any disruptions to the flow will put their passengers at risk – particularly if they are fully autonomous.

Customers know this. Sure, have fully autonomous vehicles running around test tracks or other controlled environments in  order to extend the boundaries, to generate publicity and – let’s be honest here – to have a bit of fun. Engineers are only human and enjoy their toys as much as the rest of us. (Perhaps more than the rest of us, but that is the topic of another conversation.)

Companies like Google may well find that passengers will be uncomfortable being transported in a vehicle for which there are no controls. Certainly, the first accidents that will happen on the public roads will be those caused by drivers in other cars gawking at the oddity. Therefore, due to a combination of human and physical factors removing “dumb” vehicles from certain routes or roads is the only way that fully autonomous vehicles will be able to operate safely – and that isn’t going to happen any time soon.

None of which means the industry shouldn’t be going full speed in the race to develop autonomous vehicles. As with motor racing (or the space program), it is only when technology is pushed all the way to the edge (and beyond) that real breakthroughs happen.

In the meantime, a far bigger threat to the traditional auto industry identified by Capgemini is the entry into the aftermarket business by a disruptor like Amazon. Real-time data analytics will see the aftermarket disrupted by a provider capitalizing on the increasingly connected car by showing drivers where to purchase a particular part, or delivering it in under an hour. Or directing them to a F1-style “pit stop” which undertakes repairs virtually on demand, says the company. And, as we see in this issue – many of the components can be produced through additive manufacturing (or 3D printing).

Amid all this uncertainty it takes cool heads, an intuitive understanding of the market backed by solid research and an addiction to the adrenalin rush brought on by tension to successfully steer OEMs through the multiple threats from disruptive technologies. Looking at the published results of the top (traditional) automotive assemblers it would seem that it is way too soon to confine them to the junk heap of history – as any new entrant will find to its cost. It is not only the engineers having fun testing the limits of the industry.

Previous posts

Next posts

Mon. May 27th, 2024

Share this post