More questions than answers

My mother learned to drive fairly late in life. Until I was in my teens she was happy to leave the driving to my father. But after she passed her test she said something that stuck with me: "up to now I've never really paid that much attention when someone else is driving, but now I've learnt myself I find I can never quite relax in the same way."

I don't much like being driven by other people, however proficient they are (with the honourable exception of my cousin William, who was a professional chauffeur, but he's long since retired).

Perhaps that's why I find it hard to burst with enthusiasm about driverless car technology. Now, it seems, I am badged as being anti-autonomous driving. Which I am not.

For various family reasons I have, of late, been racking up the miles on a selection of motorways, and I have to tell you that if I had a genuine, fully-functional, safety-assured wholly autonomous option on the dashboard to take me to the motorway service area closest to my chosen exit, I would select it.

I would not – as keeps being suggested – then read a book or whip out the laptop and start working because I am a lifelong martyr to travel sickness, so I would probably gaze idly out of the window until I nodded off.

What I would not do is engage my friendly autonomous system and relax if I thought there was the remotest chance of the blighter deciding it has had enough, and it was now back to me. Not least because, as noted above, I would, in all probability, be asleep.

Right from the outset of the debate about autonomous on-road technology I've been troubled by the very notion of 'conditional' autonomy. It has always seemed to me to betray a lack of understanding about the way normal people would probably behave if travelling under autonomous operation. Driving on a motorway is soporific enough already when you have control of the vehicle, asking the human driver to stay alert when not driving has never struck me as credible.

Nothing I've heard or read since then has made me any happier about this issue. I have a nagging feeling that I've got more questions than I'm hearing answers.

My starting point is that either the vehicle is capable of coping with the conditions it will encounter, or it isn't. If it can function on motorways, in mixed traffic, recognise red X signs, obey police and traffic officer instructions and cope with stretches of standing water in heavy rainfall then I'm happy to snooze all the way from home to Leicester Forest East MSA (or similar), to be nudged awake by a gentle alarm call of some sort and encouraged to pop in for a coffee. Bring it on!

The trouble is that this isn't the prospect I think I'm being offered.

Which is why we teamed up with researchers at the University of Nottingham to have a look at what might really happen when a car running autonomously decides that it wants to give control back to its unexpecting 'driver'. The results, which we will be publishing shortly, are... interesting.

Speaking of things that are interesting, I have been making my way through all 200 plus pages of the joint preliminary consultation paper on automated vehicles published by the Law Commission and the Scottish Law Commission last November. The Foundation response to the consultation will start by commending the Law Commissions for a very thorough and impressive piece of work – and they are far from done yet.

The thing that most strikes me about the consultation is the sheer number of questions that it poses – the authors claim it's 46, but in fact many of those contain four or more sub-questions. It rather bears out my theory that we are still at the stage with autonomy of having far more questions than answers.

To take just one - what, in law, do we think the term 'minimal risk condition' should be defined to mean? This being the condition that a vehicle will achieve in order to reduce the risk of a crash when a given trip cannot or should not be completed. Or, worryingly, which its human driver will achieve, having re-taken control (see above). I'm sure that through various examples this concept could be illustrated, but it's less obvious that it could ever be defined in statute (I am reminded of endless debate about interpreting the 'as low as reasonably practicable test' in health and safety legislation).

The Law Commissions are working to a three-year timescale will take them to March 2021, with a presumption that our lawmakers will pick up the baton after that. Is that soon enough for our enthusiastic international automotive businesses? Maybe not, but until we have at least as many good answers as we have questions you can count me as not anti-autonomy, just as not-yet-persuaded.