Making roads safer

How can we make roads safer? Katie Haylor spoke to transport engineering lecturer Richard Llewellyn from Edinburgh Napier University...

Richard - Some research was done many years ago looking at the causation factors behind collisions. One is that the road itself, how it's been designed, what the make-up of the road is, the way it's laid out. The second one is the vehicle itself, so how well it's maintained. And the final one is the user. What that research has shown is that 95% of all collisions happen because of user error, human error. So that really has been the focus of everything we do. And in trying to address that, historically we've had this approach called the three E's approach: one of education, getting the message across for good practice such as wearing seat belts, avoiding drink driving, and in more recent times avoiding things like use of mobile phones at the wheel.

Secondly is enforcement - so this is something that obviously the police do, it's something that is essential but it's of limited effectiveness. If we didn't have the police enforcing road safety we may well have anarchy out there, but the more we have, it doesn't have a proportionate effect. And then finally the last one is engineering, which is changing the road environment and the vehicles on it. More recently we've moved towards something called the safe systems approach, which is looking at things holistically. It's looking at those three issues, but it's also looking at what everybody can contribute to them; so it's safe vehicles, safe use, safe speed, safe roadsides, and safe post crash care; looking at the whole story of how a collision happens. And ultimately what we're aiming for is Vision Zero, where no one is killed on our roads. And basically what we're taking in as a prime assumption here is that humans are fallible. We make mistakes, and that's that's really what we're trying to deal with: minimising the damage that can be done when a human makes a mistake.

Katie - One thing that occurs to me when we talk about driver behaviour: mobile phones. Most people know that you shouldn't be holding a mobile phone and talking in it when you're driving. But some people use their phones for GPS. Some people will have a bluetooth system or loudspeaker to have a work call or a personal call whilst they're driving. What's the evidence around that? Is that any safer?

Richard - Marginally safer, but the evidence would suggest that any form of distraction within the vehicle is going to take your attention off the road. Now you wouldn't have someone in a in a factory or a nuclear power plant, for example, operating machinery while they were talking on a phone; you'd want their full attention being given to that. And when you're driving a vehicle it's exactly the same thing. And in terms of the safe systems approach, for example in terms of mobile phone use, we're now looking at taking that sort of responsibility away from the driver, but also moving it on to companies. So many companies these days, for example, now have policies that if you are driving on business you are not permitted to use your phone under any circumstances. This contributes to the safe systems approach, it's not just one person's responsibility. We all have a part to play in this.

Katie - So how do you get - just for the sake of argument - to your destination, if you're using Google Maps on your on your smartphone?

Richard - Well I mean in terms of Google Maps and sat navs, the use of audio messages through that...I mean primarily it's the conversations, the thinking on the phone, certainly operating that sat nav, making sure that that sat nav is programmed before you start that journey, making sure that it’s hands free; and also things like, you know, you take regular breaks.

Katie - We live in an era that is ushering towards autonomous cars. We're not really there yet, but what we do see is increasing levels of autonomy in our vehicles. How much of that can we expect in the near future then, and is that going to make a difference to road safety?

Richard - Well the most immediate actions that are being taken were outlined by the EU a couple of months ago. There was a political agreement in March of this year to add a couple of key things to vehicles that are produced and used in the European Union from 2022. And one of those is something called intelligent speed assistance. It works using a combination of GPS, cameras, radar technology to determine the position of the vehicle; and it compares that with a speed limit map. Now if the vehicle finds that you are driving in excess of the speed limit, there's a couple of options. It will either advise you you're doing so, or in ultimate circumstances it may even forcibly reduce your speed. Another measure coming about, and in fact is is fitted on quite a few vehicles already, is this concept of an emergency braking system. This is a system that detects objects such as a vehicle in the road ahead. And if you as the driver don't apply your brakes in time, these systems will automatically bring you to a stop. So these types of things have great potential to really make a difference in terms of the number of collisions on our roads. But there is also potentially a negative side with these things. In the rail and air industry, for example, we found when automation is taking place to vehicles, actually drivers’ attention or pilots’ attention starts to wander as the mind has less to do; that can create other problems. So it's not by any means perfect but hopefully a step in the right direction.

Katie - Is there another issue here in that these new cars that are coming in have these added safety features - that's great - but if the person behind me that goes into the back of me is in a really old model that isn't particularly autonomous, I guess there's still a risk there, is there?

Richard - And this is one of the big challenges I think over the next few decades of autonomous vehicles and how they're introduced, in a situation where an autonomous vehicle is completely free to do what it wants. We've got great technology: for example, Stanford University has just developed a race car that can be driven around a race track at speeds commensurate with that of a racing driver. It can deal with all sorts of conditions, wet conditions, and it's got really, really good detection systems. But you put that in a real environment where it's mixed with other vehicles, and indeed human beings making decisions on that road, and you start to have a problem.

Motorways for the most part are reasonably predictable, and I think in the future that's probably the first place that we will start to see autonomous vehicle use really starting to take off, perhaps we might see dedicated lanes to start with. When we move into the urban area things get incredibly complex. Humans are unpredictable. And going back to that fact that 95 percent of collisions are due to human error, those vehicles start having interactions, which really any amount of artificial intelligence is really going to struggle with. So dealing with pedestrians walking out into the road, interactions...I mean if you or I were at a junction and someone flashed our lights to let someone out, we'd know exactly what to do. But what would an autonomous vehicle do? And there's also a lot of other questions. A moral question: if an autonomous vehicle was in the situation where there were other vehicles and other users on the road, and it had to take evasive action, what does it do? And who would be legally responsible? And there's other issues like cybersecurity, people starting to hack into vehicles. So there's a lot of questions that we really need to grapple with in the future, and arguably these moral and legal questions are just as tricky as some of the technological ones.