Dr Shaun Helman, Chief Scientist for Transportation at the TRL discusses the impact of new vehicle technologies on driver training with the Intelligent Instructor.

Over the last decade my work at TRL has afforded me plenty of opportunities to listen to the questions people ask at conferences concerned with driver training and testing. One of the more common is similar to the one I have been asked to address in this piece. It is along the lines of “Do we need to modernise driver training and testing to bring them up to date with new vehicle technologies?”

In order to approach this, let’s consider the technologies. First the existing systems, then the near future, and then the (slightly) more distant future.

Consideration of just the existing situation suggests that there may indeed be a need for us to update what we do. There are existing systems which, arguably, are not properly covered in testing or training. For example, Advanced Braking Systems (ABS) have been required on all new passenger cars sold in the EU since 2004. Many driving instructors I have spoken to over the years have expressed their surprise that drivers are not always taught to use these properly (for example by steering while braking, which is one of the advantages of such systems, or even simply understanding how the system feels under heavy braking).

A future near you

A good source of information regarding which technologies are likely to appear as standard on new vehicles relatively soon is the extensive report produced by TRL for the European Commission (Seidl, Hynd et al., 2017). This considered 24 different vehicle technologies overall, including 19 which might be considered as standard for cars sold in the European Union (and presumably the UK, Brexit notwithstanding). A number of these technologies are arguably dependent on the driver’s ability to understand and use them (or at least in some cases make sure they are switched on); Autonomous Emergency Braking (AEB) and the same system for preventing collisions with pedestrians and cyclists (AEB-PCD), Intelligent Speed Assistance (ISA) and Lane Keep Assist (LKA).

Down the road

‘Driving’ will change fundamentally with the advent of autonomous systems taking over the driving task itself, at least on certain roads. If the major market players are right in their optimism around such technology, this is closer than we might think. Ford, for example, is on record as saying that it intends to bring a fully autonomous vehicle into the market by 2021. Other companies are talking about similar timelines. It is, of course, likely to be much longer than this for such technology to become ‘the norm’, but even so there is a great deal of research going on in this area. Much of this is focused on the best ways to signal to drivers the need to take back control of driving when such systems reach the limits of their abilities. There will also be basic human factors issues to solve: the shapes of the buttons, the tone of the warning signals, and the way in which the system provides feedback to the driver – all of these things will have major implications for how effectively they work, and how we might train people to use them.

All change

In short then, the answer is unsurprising. Of course we need to adapt our training and testing to account for new technologies. What might be more surprising is that we have been doing this all along. You would not believe this from the way the question is asked at conferences; what I often detect in the tone of people asking it is an implication that the driving test, and driver training, have not really moved with the times at all. This is demonstrably not true. For example, a glance at the excellent ‘history of road safety, the highway code and the driving test’ page at the www.gov.uk website reveals at least a couple of dozen changes to the practical car driving test (excluding price increases) since it was introduced as a mandatory test in 1935. These include obvious responses to technological developments, such as the introduction of a separate licence group in 1969 for vehicles with automatic gearboxes, candidates no longer having to display arm signals in the test from 1975, the introduction of ‘show me’ and ‘tell me’ questions in 2003, and assessment of eco-safe driving from 2008. Change may have sometimes been slow, but it has still happened.

The most recent change to the test (expansion of the independent driving section to allow use of a SatNav in 2017) also has a technology focus to it. However, along with the introduction of independent driving by following road signs in 2010, on the basis of the research TRL undertook to underpin both these changes I would argue that they are simply to do with the general notion of aligning the test to the driving tasks people engage in after they pass, rather than just being about technology. In short, the more like ‘real driving’ we can make the test the more valid it should be as a measure of the competence of drivers to actually use their vehicles appropriately. Obviously, if the driving task changes (as it has in the past with the introduction of automatic gearboxes, or no longer having the need to use arm signals) we need to change the way we train people, and test their ability to implement this training.

I would argue that this basic problem is solvable, even if we stick religiously to the current model of driving lessons followed by a pass/fail test (more on this below). We already have a test that can adequately include a range of technology use cases to check for competence. Consider the questions from the ‘show me’ section of the test. “When it is safe to do so, can you show me how to wash and clean the front windscreen?” is not a fundamentally different question to “…can you show me how you’d turn on your AEB system?” or even “…can you show me how you’d take back control from your autonomous vehicle?” All of these things would potentially be tested in periods of independent driving anyway; just as now, some unlucky test candidates have to endure scrutiny of their skills in wet weather, in the future, autonomous systems will want to hand back control to candidates at inopportune times in the middle of their test.

Zebra crossings

Of course, there is detail to be worked out. One particularly important issue (non- standardisation of systems) might be best illustrated by the following anecdote. The day I was asked if I would like to produce this article, I had a hire car delivered for a journey I was undertaking the following morning. After the car was delivered, I did what I always do; I sat in it and asked myself the following six questions:

  1.  How do I switch the car on and off?
  2. How do I put the handbrake on and off?
  3. How do I open and close the fuel cap?
  4. How do I turn off or at least reduce the brightness on the central console screen to avoid it distracting me in my peripheral vision when driving?
  5. How do I connect my phone to the in-car entertainment system so that I can listen to music on my journey?
  6. How do I turn off traffic announcements and the various warnings modern vehicles offer about things like speed cameras (so that said music listening is uninterrupted)?

On this occasion, it took me around two minutes to do all but one of these; I needed a more skilled colleague on the journey the next day to Google the answer to number 6, after traffic announcements had interrupted our musical enjoyment four times. (As an unrelated observation, I can report that the second side of Abbey Road, considered by many as the pinnacle of pop music, sounds considerably better without continued interruption by radio announcers discussing traffic flow between Junctions 8 and 10 on the M40).

I digress. The point is that the time required for my vehicle familiarisation should be both unsurprising and surprising. On the one hand, car interiors vary greatly in their design, and it is only to be expected that this variation necessitates some preparation every time a new make or model is encountered. On the other hand, cars are the most dangerous machines most people use in their lifetime, and one might expect that a rational approach to this would be to introduce as much standardisation as possible into their design. It is clear that the existing approach to standardisation and guidance for design is failing to achieve the ‘utopia’ of usability – everyone being able to understand instantly how to use a given technology, without needing to consult a manual or the internet.

The impact of variation in design on how we train people to use features in vehicles is obvious, especially for people who do not present for the test in the same vehicle in which they have been trained; a ‘show me’ question on a given feature may lead to failure if the feature being tested has a completely different design to that on which people have trained, and is not intuitive.

Why do that?

That is not the end of the story though. Note that my conclusion above (‘the problem is solvable’) was accompanied by the caveat “...even if we stick religiously to the current model of driving lessons followed by a pass/fail test”. I would argue that there is another ‘advancement’ which has so far proven NOT to be solvable (at least not completely) with respect to driver training and testing, but which CAN be solved through technology. This is our advancement in understanding of driver behaviour (as distinct from driver performance).

Evans (2004) defines the latter as what a driver CAN do, and the former as what as driver DOES do; a driver with a high degree of cornering skill (performance) can nonetheless crash because they are overconfident and decide to choose much too high a speed into a bend (behaviour). Evans points out that while performance can be measured and tested relatively easily and reliably, the same is not always true of behaviour, due at least in part to the fact that people can ‘fake it’ when being tested.

Consider the existing situation. The group which most readily passes the practical driving test (young males) is also the group which most readily gets involved in collisions involving death and catastrophic injury. Individual cases will vary of course, and the test clearly sets some kind of minimum standard, but as a public health intervention (which given what we know about the risks involved in early driving, it surely is) the test does not seem to achieve its goal as well as it might.

Interestingly, there is a technology angle here too. The number of ways in which we can gather driver behaviour data from people (through telematics devices, smartphone apps, cameras and so on) has grown a great deal in the last decade. What if this technology is also considered in the remit of our original question? What if we didn’t have a single test in which your ability to ‘do the basics’ was tested, and in which your behaviour could be artificially controlled through your own desire to be seen in the best light possible? What if instead we had a first test that took place over a single session to measure performance, and then a second test which took place over an extended period of random remote monitoring to measure behaviour? What if you  knew that in your first 2,000 miles of post-test driving, a random 200 miles (on different types of road, at different times of the day, and in different driving conditions) would be monitored automatically and in great detail for how you were choosing to behave? What if, on the basis of the way you were choosing to behave, you could lose your licence? This is what the telematics insurance market claims to be trying to achieve, but built into the licensing system and without the encumbrance of a risk portfolio to be managed.

Tomorrow assured

The most important implication for new vehicle technology on training might not be the obvious - needing to ensure that people are trained and tested appropriately on their ability to use new vehicle technologies - but actually the other way around; using new vehicle technologies to ensure that driver training and testing actually focus on how people behave when they are engaged in ‘real world’ driving, as well as when they know they are under scrutiny. This would be an opportunity to not only include new technology in the current training and testing paradigm, but to use it to define a new one.

View this month's magazine issue

Register for our newsletter

Learn more about TRL with regular updates

Media membership

Our latest press releases and news

Get in touch

Have a question? Speak to one of our experts today.