By Professor Stephen Glaister, director of the RAC Foundation.
Like many people, I have watched the news clips of Google’s self-driving car in the US with amazement that it can be done to a standard that allows it onto a public road. But also with scepticism that it would be sufficiently safe in the real world and that it will ever be cheap enough to become a sensible proposition for the ordinary car buyer.
Recently the BBC’s Richard Wescott has reported on a less sophisticated but cheaper system at the University of Oxford. It is still far too expensive to justify replacing tasks that we actually find reasonably undemanding to do for ourselves—like driving to our regular place of work. But if, and when, it becomes cheap enough people will start to buy something of this kind.
It is easy to dismiss these ideas on the grounds that they will never be safe enough in the rough and tumble of the typical, busy high street. Of course, safety will have to be properly proven. But we must not forget that human drivers are not entirely safe. They – we – make mistakes. Official Department for Transport figures show that ‘driver/rider error or reaction’ was a contributory factor in 67% of fatal accidents in 2011.
Even highly-trained commercial airline pilots make errors and modern aircraft are full of electronic devices to detect and prevent or correct pilot error—or to take the human out of the system altogether. Planes can now navigate themselves and land fully automatically even at airstrips that are not equipped with instrument landing systems. The general public driving in the chaos that is the high street can benefit as much or more from electronic assistance.
We tend to forget the extent to which electronics have already become commonplace in our cars. We take for granted engine management systems, satellite navigation and parking aids. Then there are anti-lock braking systems (ABS) and electronic stability control (ESC) which do a far better job of keeping a car under control in extreme circumstances than the ordinary driver could do. These have already saved many lives. Now we are getting continuous tyre pressure sensing, detection of threatening approaching vehicles at junctions and systems to implement emergency braking when a crash is inevitable. Automatic speed control is already fitted to some vehicles.
A family car just launched at an ordinary sort of price offers the following as standard: Active Cylinder Technology to improve fuel consumption, electronic parking brake, navigation systems, automatic distance control including Front Assist with City Emergency Braking, XDS electronic differential lock, electronic tyre pressure monitoring system, a lane assist camera, energy recovery to a battery during braking, front and rear parking sensors, multiple impact brake activation to reduce the chance of a second impact.
So much has changed in the last few years: people do buy all this electronic assistance if it is perceived to be good value for money. They accept it as safe—they know the liabilities the manufacturers would face if it were to transpire that they were at fault. With electronics much of the cost is in the research and development of hardware and software. Once that is done manufacturing costs can be low. So it seems that the self-driving car at an acceptable price may not be so far away after all. It will probably creep up on us before we have noticed.
Has there been any research on whether systems like cruise control can cause accidents?
Some electronics (e.g. ABS, rear parking sensors) improve driving abilities and are relatively uncontroversial. Others such as cruise control or lane monitoring can take over some driving functions – possibly causing drivers to drive with less attention. In these situations full “auto-pilot” may indeed be safer than the partial auto-pilot that is “creeping up on us”.
Unfortunately there is a large cost and technology gap between partial and full auto-pilot. The latter may also require all vehicles to carry some equipment and for all(?!) roads to have some level of technology (a minimum of reflective kerb, lane and junction marking).
If the government is going to struggle to get us to replace our FM car radios with DAB radios, I don’t think they stand a hope of compelling us to install auto-pilot type technology.
I have feeling there have been some press reports of the technology failing and throttles getting stuck, but are rare occurrences. As is often the case technological advances come quicker than changes in public attitudes. There will also be those who say driving is not just a chore but something to be enjoyed. Fair enough, but at what cost to other members of society?
I was not thinking of throttle cables getting stuck on cruise control equipped cars (they jam on older cars as well). I am resonably confident (despite a few scare stories) that car manufacturers will get the technology reliable – as you say “liability” will be a great moderator on their ambitions. It’s that bit of software in the driving seat and nominally responsible for the vehicle that is the worry – it has not really evolved very much since we started riding horses.
My concern is that if you go on cruise control (say 68mph), get in lane 2, switch on that little gadget that keeps you in lane, you can relax and possibly pay less attention to the road – after all speed and direction are now under control and people should not pull out in front of you nor cut in in front of you …
I suspect that there have been a number of accidents (such as rear-end shunts or collisions due to not responding to someone else changing lanes) because the driver is on “auto-pilot”. However, I have never seen any statistics – I doubt accident investigators are able to check whether features like cruise control were on immediately prior to a collision. Perhaps we can see whether cars with cruise control etc. fitted have more of a certain type of accident than similar cars without.
If people genuinely become less attentive as technology takes over, you need the more advanced technology as well – such as automatic distance control, safe-zone scanning and inter-vehicle signalling – plus an alarm to wake the driver when approaching his exit.