Human error is a frequent contributor to road traffic collisions, understanding the human factors provides a deeper insight into road-user behaviour.

Call us on 0845 619 1275

Archive for December, 2010

Do you know where you’re going? The psychology of satellite navigation system use.

Friday, December 10th, 2010

Over the past ten years or so there has been a massive increase in the amount of technology that we can use when driving. While this technology can either entertain us or aid our driving, it can also distract us. In fact, some research has suggested that distractions may account for up to 78% of crashes[1]. Much of the research and media attention has been focused on the use of mobile phones when driving; however the use of satellite navigation systems has become commonplace, with over 14 million in use in 2008.  Relatively little is known about the advantages and disadvantages of using these systems.

 

iStock_000002494679Large The main problem with using mobile phones when driving is that your attention is divided between two tasks: talking and driving. You cannot give either your full attention, therefore your driving skills decrease. There seems to be a similar effect when using sat navs. Cases of drivers ending up on the edge of a cliff, on railway lines or on roads too narrow for a car to pass through have been well publicised in the media. If they were paying attention how could this have happened? What has research shown us about how we use satellite navigation systems?

Intuitively, it might seem that the use of a navigation system would be less distracting than using a mobile phone. However, research has shown that drivers are more likely to be distracted by a task that is relevant to their driving, such as map reading, than to something that is irrelevant[2]. It’s therefore possible that the use of sat navs might distract a driver or interfere with their ability to drive safely.

While there may be some negative effects from using sat navs, they are used for a good reason – to navigate our way to a destination. Drivers have needed to achieve this since long before sat navs came into use! Maybe a more relevant question is whether sat navs are safer, or less distracting, than the traditional paper map?

One study compared driving performance when using either a sat nav system or a paper map[3]. They found that drivers using a sat nav system reached their destination faster and took a shorter route to their destination. However, the sat nav users also drove significantly faster and drove more aggressively around corners. So it seems that the use of sat navs in fleet cars would bring some advantages in terms of efficiency, but may bring some increased risks at the same time.

It’s unlikely that the use of sat navs will decrease and they have many obvious benefits for drivers, so how can we minimise the negative effects of using sat navs? There seems to be two approaches to this issue.

First, much research is being conducted to improve the sat nav systems in order to reduce the cognitive load that is placed on the driver, so that they can focus their attention more on driving. One line of research is comparing the usual head down display with a head up display for the use of sat nav systems when driving. One study asked commercial lorry drivers to drive in a simulator using either a head down or head up display[4]. They found that drivers using a head up display responded faster to emergency situations and maintained more consistent driving speeds.

Current sat nav systems provide the driver with information about when to complete a manoeuvre on the basis of distance. One research group[5] is examining how verbally identified landmarks, for example “turn right after the pedestrian crossing”, might be used to improve the information that drivers are given by navigation systems. Half of the participants received direction on the basis of distance and half on the basis of landmarks. They found that the participants being given landmark instructions were more confident in their driving, spent less time looking at the sat nav display and made fewer driving and navigation errors.

The second approach is to help drivers to better understand the risks involved when using a sat nav and to develop driving techniques to avoid the possible dangers that are encountered. For example, being more selective about when to look at the device, not making sudden changes of direction and trusting your own common sense, can help you to drive safely whilst still having the advantages of a navigation system.


[1] Neale et al. (2005)

[2] Cnossen et al. (2004)

[3] Lee and Cheng (2008)

[4] Liu and Wen (2004)

[5] May et al. (2005)

Dr Victoria Bourne  (BA Hons, DPhil)

Consultant to Driving Risk Management Limited

Looked But Failed To See Errors

Saturday, December 4th, 2010

After an accident, people often report that they failed to see the other vehicle involved in the crash with enough time to avoid the collision, even though they looked in that direction. These “looked but failed to see” errors are very common and it has been estimated that not seeing another vehicle is implicated in up to 50% of collisions. Why are we so bad at detecting other vehicles, even when we actively look for them?

Look but fail to see errors can largely be explained in terms of the way in which humans process information. Our subjective view of the world feels rather like we are watching a film and that we smoothly look around the world with a perfect representation of the outside world being efficiently processed by our brain. Surprisingly, this is very far from the truth.

When we look at something, we actually only have clear vision in a small, central part of our visual field. If you hold you thumb up in front of you at arms length, our clear patch of vision is about the size of your thumbnail. Outside of this patch our vision becomes increasingly blurry. In addition to this, the way in which we move our eyes is not at all like the smooth camera we perceive. We tend to fixate our eyes on something for around a quarter to a third of a second and then move our eyes to a new location with this movement, or saccade, taking around 30-50 milliseconds. When we are moving our eyes we are, effectively, blind.

So, our visual input is actually very degraded. Each second we will see about three “snapshots” of our visual world and these snapshots are mainly quite blurry. In reality our vision is more like a cartoon flip book where a slightly different image is shown on each page, but flipping quickly through them gives the impression that the characters are moving. In addition to being blind when our eyes move, we also have no visual input when we blink. These two factors combined mean that we are, essentially, blind nearly 20% of the time! Additionally, research has shown that drivers at a junction may only look at the road that they are merging with for less than half a second[1].

Why does it seem like we have a clear and smooth moving perception of the world, when actually we only have blurry snapshots? Essentially our brains do some amazing “piecing together” of our snapshots and allow us to perceive a clear and smooth moving world. Part of this filling in is done on the basis of our experiences and existing knowledge. Imagine driving up to a busy junction near to where you live that you cross everyday. Your memory of this junction, called a schema, is part of what allows your brain to create a complete and clear image of it. Given this it’s easy to imagine how looked but failed to see errors can occur.

The limitations of our visual processing system can explain, at least to some extent, look but failed to see errors. But other factors also contribute. How conspicuous, or noticeable, a vehicle is also makes a big difference as to whether a driver sees it or not. Much of the research in this area has concentrated on motorcycle conspicuity given the relatively high frequency with which motorcycles are not seen before a collision. Given that the front of a motorcycle is relatively small, it seems quite obvious that it would be easier to miss, but what can be done to reduce this risk.

The use of headlights and wearing of bright clothing have been examined as two of the most likely ways of making motorcyclists more visible. The use of headlights during the daytime has been found to reduce the number of motorcycle accidents; however, interestingly this is also true for cars, trucks and buses[2]. It has also been suggested that headlight use is only increases conspicuity when a motorcycle is further away and that it is little use at shorter distances[3].

Research has also examined whether there are differences in looked but failed to see errors between novice and expert drivers. Typically these studies monitor where the driver is looking and then these patterns are compared between the two groups[4]. The experienced drivers had very different looking “strategies” when driving on a motorway, urban or rural road. This suggests that they vary their driving techniques according to the environment. In contrast, novice drivers had very similar patterns of eye movements and fixations on all three of the road types. One implication of the experienced drivers being able to adopt suitable driving behaviours for different road types, this may also come at a cost. Because of all their experience, these drivers are likely to have well established schemas of the roads they frequently drive on. Consequently, they may be less likely to see and oncoming vehicle if it does not “match” with their existing schema. Ironically, this means that experienced drivers may be more likely to commit looked but failed to see errors than novice drivers.

Although looked but failed to see errors are relatively common, there are techniques that can be used to minimise these errors occurring, and consequently the associated risks. Observation is a key skill for driving, but one that is often overlooked and taken for granted. Training can improve observation skills through developing visual search strategies.  Such methods can reduce the occurrence and risks that arise due to looked but failed to see errors.


[1] Langham (1999)

[2] Olson (1989)

[3] Hole et al. (1996)

[4] Underwood

Dr Victoria Bourne  (BA Hons, DPhil)

Consultant to Driving Risk Management Limited

Driving when tired – how and why does fatigue influence driving behaviours?

Friday, December 3rd, 2010

Driving when feeling very tired is obviously highly risky behaviour and estimates have suggested that up to 20% of collisions could, at least to some extent, be attributed to fatigue. So how does being tired influence our driving behaviours and what can we do to reduce or avoid our chances of being involved in an incident?

One of the ways in which fatigue impairs driving behaviour is by making reaction times slower. In one study[i] 240 male drivers were stopped between 1am and 6am. They were asked to rate how tired they were and to complete a simple reaction time task. Those who felt rested had reaction times of around 189 milliseconds (ms), but those who were very tired had significantly slower reaction times of 309 ms.

However, some research[ii] has suggested that this increase in reaction times might not be seen in actual driving behaviours. Instead, it is argued that, when fatigued and driving, we allocate our resources in such a way that we prioritise collision avoidance, but that other tasks, such as steering accuracy, may suffer. There is also some evidence to show that driving performance when fatigued is more affected when driving an easier route than when driving on a more complex one[iii]. It seems that the added cognitive load enables the driver to allocate their processing resources in a way that leads to safer driving behaviour.

Regardless of why fatigue influences driving behaviour, it is clear that the chances of having a crash increases greatly when the driver is tired. The important issue is how this risk can be avoided or minimised. Two psychologists, Reyner and Horne[iv], have conducted a series of experiments examining how effective various methods of reducing fatigue are.  Methods such as listening to loud music or opening the window are only effective for a very short period of time, around 15 minutes. Similarly, consuming a highly caffeinated drink only relieved sleepiness for about 30 minutes in people who had no sleep at all the night before, although it was more effective if people had at least some sleep. The researchers concluded that by far the most effective method of avoiding fatigue was to take a short nap, of just 10 to15 minutes, and then have a caffeinated drink.

Some people are more at risk of driving when fatigued than others, and those who drive as part of their occupation are particularly vulnerable. There is evidence that the attitudes of employers can influence how well drivers deal with fatigue. One piece of research[v] examined how the organisational safety climate and occupational stress might influence driving behaviours whilst fatigued. They found that fewer “near misses” occurred when drivers exhibited lower levels of work-related stress and worked for organisations with a high awareness of driver safety-related issues. However, while both factors were found to be significant and important, the occupational safety climate had a far greater effect. This study really emphasises the importance of the attitudes, policy and training provided by employers in reducing potential on-road incidents.

For instance, research has proved conclusively that there are two times of the day during which collisions due to fatigue are most likely to occur: at night time between midnight and 6 a.m. and at mid-afternoon in the post-lunch dip. A sensible risk management action would therefore be to arrange work patterns so that no business driving was required in the early hours and that drivers were mandated to take a break during the afternoon period.

Ideally, people would not drive when they feel sleepy, but in the real world this is hard to police and prevent. One thing that is clear from the research is that the risk of having a crash can be reduced if driver safety is taken seriously by both employer and employee. This involves having policies in place, procedures to identify likely risk scenarios and interventions that are tailored to specific need.


[i] Corfitsen (1999)

[ii] Van der Hulst et al. (2001)

[iii] Matthews and Desmond (2002)

[iv] Reyner and Horne (1998, 2000), Horne and Reyner (1996, 1997)

[v] Strahan et al. (2008)