# C.0018

## Driving

During the early days of motoring, cars were unreliable, expensive, and difficult to drive. People bought them partly because of the challenge - motoring was a sport, and curiously, we still buy machines that are called ‘sports cars’ even though we have no intention of racing them. The fact is that many of us find cars fascinating. They give us privacy, social status, and an outlet for physical expression [2] [5] [11]. The fascination starts at an early age. As children, we enjoy sitting in the driving seat and holding the steering wheel. As teenagers, we long for the keys to the family saloon. In more than one sense, driving means being in control, and the way we think about cars and the way we think about other road users make driving behaviour a rich field for psychological study [8]. But here, we shall be concerned with the driving process at a rather basic level, in particular, the visual environment as perceived from a moving vehicle, the physical interface between driver and machine, and the routine operations a driver must carry out to get the vehicle from A to B.

### The visual field

Travelling in a car changes the way we see our surroundings. Like most animals, we have evolved to deal with scenery that stays put because most of the time, we’re standing still or nearly so. The distant mountains remain distant. When we walk down the street, houses roll into view at a leisurely pace. It’s true that the act of turning one’s head causes distant objects to move across one’s field of vision. If you swivel round 180 degrees, a distant hilltop will appear to travel several kilometres round a semi-circle at impossible speed. But your brain doesn’t take this to mean that the world is spinning around you, because it has evidence from other sources that your head has moved.

Now picture yourself in a car travelling at 100 km/h along a country road. When the car travels round a curve, trees on the horizon move across your field of view, but this time you haven’t moved any part of your body at all. This is a new experience, and the brain must re-interpret what is going on. Even on a straight road, the visual field is no longer static: everywhere you look, the landscape seems to be moving, except for a single point directly ahead at infinity. Everything else appears to expand from that point (figure 1). Nearby features move past you more quickly than the distant landscape. The closer the object, the more dramatic the impression of speed.

This has interesting consequences on wide roads such as motorways. Compared with ordinary roads, the verges and hedgerows that tell us how fast we are moving are further away from our line of vision, and often the nearest feature will be a platoon of vehicles in a neighbouring lane. The platoon provides a new frame of reference that is moving along the road with us. In this new frame of reference, we are standing still, or if the neighbouring cars are moving faster, we can even appear to be moving backwards. So it is not surprising that in foggy weather, when the cues provided by roadside features disappear, drivers both underestimate their speed and drive too close to the vehicle in front [6].

There are other cues that consciously or subconsciously influence our perception of speed. One is the noise that comes from the engine, tyres and transmission, together with the aerodynamic wind noise that penetrates the body shell. Another is vibration transmitted from the road surface through the tyres and suspension. Inevitably, noise and vibration are diminishing from one generation to the next as cars become more sophisticated. It has been suggested that a smooth, quiet ride insulates the driver from the road environment and reduces awareness not only of speed but also of road hazards such as a bend in the road ahead [10]. Nevertheless, drivers in many countries have fewer crashes than they used to, in the sense that the injury rate per kilometre travelled continues to fall year by year. Perhaps if cars were noisier and less comfortable, it would have fallen further.

Driving can be broken down into different types of control task. Some are pre-planned, and some are continuous. A pre-planned manoeuvre is one in which the driver conceives in advance the actions to be carried out. One example would be turning into a side road from a main road. Having gauged the road layout, the driver performs the actions in quick succession. They include checking the mirror, looking out for pedestrians and other vehicles, braking, changing gear, and turning the handwheel. This is not to say that the driver could finish the manoeuvre blindfold: at each stage, corrections will be needed based on feedback concerning the car’s position and speed. But thinking ahead allows the driver to accomplish several tasks in quick succession that would otherwise be hard to manage. Put the other way round, if the plan is flawed, disaster will follow.

Continuous processes are different because very little planning is needed. For example, on a quiet rural road the driver can maintain a more-or-less constant speed and lane position by making small adjustments whenever the need arises. These two different types of driving task seem to require different skills and different levels of concentration. We’ll consider them in reverse order, starting with some periodic adjustment processes first, and going on to examine some specific manoeuvres later.

In practice, one suspects that drivers don’t think in terms of distances or times, but rather continue to reduce the gap until the vehicle ahead looms uncomfortably large within the field of view. Let us put this aside, however, and assume there is a target position and consider how the driver might move towards it (figure 2). This is unlikely to be a pre-planned manoeuvre, for two reasons. Firstly, human drivers can’t judge distance gaps or time gaps very accurately. Secondly, for reasons that we shall explain later, neither the brake pedal nor the accelerator pedal of a normal car constitutes a precise method of control. Whatever your target, you will converge on it by applying a series of corrections. If you seem to be lagging behind, you press the accelerator for a while. If you seem to be too close, you lift your foot off the accelerator for a while, and possibly switch to the brake pedal. The whole process can be pictured as a feedback loop in which your spacing, like the temperature in a thermostatically controlled hot water system, zigzags between acceptable limits.

As well as avoiding the vehicle in front, you will also want to keep your car in the correct lane. Picture yourself driving along a two-way rural road consisting of a series of shallow curves and the occasional straight section. If the car strays towards the middle of the road, you apply a correction to the handwheel. If it gets too close to the kerb, you apply a correction in the opposite direction (figure 3). In this fashion you can cover hundreds of kilometres on a motorway without consciously making any judgements about the road layout at all: you perform a tracking process similar to the car-following process described above except that the adjustments take place from side to side and tend to be more precise. Steering is easier to judge because the target distances are smaller, and the steering controls of a family car allow most drivers to position their vehicles within a few centimetres of their desired path.

#### Cornering

Tracking processes represent a particular kind of control behaviour at one end of a spectrum. Pre-planned manoeuvres represent processes at the opposite end of the spectrum. Somewhere in the middle lies the process that racing drivers and motoring journalists call ‘cornering’. We have already touched on the behaviour of a car when it follows a steadily curving path of constant radius in sections C0415 and C0418, and now it’s appropriate to say a little more about the transitional behaviour that occurs as the vehicle switches between straight and curved paths at the entry and exit where the dynamics of the body mass and suspension are continuously changing.

One can only guess how drivers cope with the dynamics of vehicle motion during these transitional phases, although there are several cues they could use to assess what is happening. One specialist [1] has drawn attention to yaw velocity and lateral acceleration. Yaw is just a change in the direction you are pointing, and you will know it’s happening when distant landmarks appear to move across your field of vision – the world is slowly spinning around you, and yaw velocity measures the rate at which this happens. Lateral acceleration is perceived quite differently, as sideways pressure exerted by the driving seat on your trunk and thigh.

Both yaw and lateral acceleration occur when cornering but they don’t necessarily occur at the same time or change at precisely the same rate. Imagine you are walking along a straight line painted down the centre of a shopping mall. Now ‘yaw’ a little to the left, turning your body to look at a particularly interesting café as you pass by. You can do this without deviating from your course – there is no lateral acceleration because you are not moving in the direction you are pointing. Although less flexible than the human body, a car can do this too. Hence yaw alone will not get your car round a corner. It needs lateral acceleration as well, delivered through forces acting sideways on the tyres, and these take a moment or two to build up.

Now imagine you are in the driving seat. After travelling along a straight road for a while, you are approaching a corner. As you turn the handwheel, the first thing you’ll notice is yaw velocity. The horizon moves across the front of the vehicle as the nose of the car starts to swing across. Next, you feel lateral acceleration as the slip angle of the tyres builds up (see Section C1717), developing a sideways force that deflects the car away from its original (straight line) trajectory and guiding it round the curve. Both act as cues from which you judge the severity of the bend and how much grip is left in reserve, but they don’t occur simultaneously. The reason is the lag between yaw velocity and lateral acceleration. Once you are into the part of the bend where the curvature is more-or-less constant, things settle down. It’s the transient conditions at the beginning and end of the curve where the cues are out of phase.

#### Lane change

Cues can actually contradict one another when the driver tackles two opposite-handed curves in quick succession, as does a racing driver when entering a chicane on a racing circuit. Ordinary drivers undergo a similar experience in less extreme form when switching between lanes on a wide road. This can pose a significant challenge if it has to be carried out as an evasive manoeuvre. Imagine you are driving along a motorway and need to move to the neighbouring lane on your left. You turn the handwheel, the vehicle yaws to the left, and then begins to move across. You feel a lateral acceleration against your right hip and shoulder. Halfway through the manoeuvre, you begin to rotate the handwheel in the opposite direction. But owing to the complex transient behaviour of the vehicle mass and suspension, the lateral acceleration continues until after you have ‘unwound’ the steering and after the vehicle starts to yaw to the right. At this point, the two cues are delivering contradictory messages [1]. It seems that drivers find this disconcerting, and perhaps explains why ‘slalom’ manoeuvres are given such high priority in vehicle testing programmes: factory test drivers push new models to their limits in order to expose any quirks in behaviour that might take customers by surprise. But a car can never be entirely foolproof, as demonstrated by the widely publicised roll-over on which two Swedish journalists were injured while testing the new Mercedes-Benz A-class vehicle just before launch in 1997.

### The control interface

A road-going vehicle can be manoeuvred independently in two ways: in the fore-and-aft direction, and from side to side. In the first case, we use accelerator pedal and brake to control the motion, and in the second case, the handwheel (steering wheel). But the control process is not as direct as you might think, because there are several intermediate stages before a control action takes effect.

Some readers may remember an early computer game in which two players were each equipped with a joystick. One joystick controlled a paddle that moved up and down the left-hand side of the screen, while the other controlled a paddle that moved up and down the right-hand side. A virtual tennis ball flew back and forth across the screen between the two. In one version of this game, an interesting feature of the controls was that the position of each joystick was mirrored immediately by the position of its paddle. If you moved, say the right-hand joystick away from you into a new position, the paddle moved straight away to a corresponding position on the screen. If you moved the joystick back to its central position, the paddle did likewise.

This is easy to arrange with the appropriate software, because a computer game deals only with electronically-created images. But real objects such as animals and vehicles cannot move instantly to a new position because forces must be applied to overcome their inertia. That is why the controls of a vehicle don’t cause it to move instantaneously from position A to position B. Instead, they alter the levels of power that are fed into the actuators that cause the movement; in other words, they control acceleration.

#### Brakes and accelerator

Take, for example, braking to a halt at a traffic signal. You cannot instruct the braking system in advance, as it were, to halt at the stop-line. The only thing you can do is press on the brake pedal to produce a greater or lesser degree of (negative) acceleration. The speed falls at a rate proportional to the acceleration, and the vehicle only comes to a halt when the speed finally reaches zero. Exactly whereabouts along the road this will happen is hard to judge in advance. The braking system does not permit precise control.

The same is true of the accelerator. If anything, the connection here is even less direct. Pressing on the accelerator does not move the vehicle instantaneously to a target position say 500 m further along the road. It merely adjusts the engine fuel supply, which in turn (possibly after a couple of gear shifts followed by an exchange of angular momentum between the engine, clutch, gearbox and the transmission) determines the power output and therefore the rate of acceleration.

It follows that the progress of a vehicle along the road is only tenuously connected to the pedal positions. Quite large pedal movements will take place during the course of a journey that are not reflected in the vehicle’s position nor even to a great extent in its speed, and this has been demonstrated by recording the pedal movements and the vehicle speed over test drives on public roads. This is one of the reasons why the movement of dense streams of vehicular traffic on motorways is unstable. Vehicle spacings are erratic.

#### Instrument panel

Keeping a safe gap is a far-from-trivial task, but during the early days of motoring, there were many others. The driver would fuss with auxiliary controls such as the ignition timing and ‘choke’, and changing gear called for a special technique. A drive in the country was an adventure, and arriving safely at one’s destination an achievement. By contrast, the 21st century motor car can be turned on and off like a tap. When we ride in one, we are barely conscious of what goes on within the engine compartment. Especially in a car with automatic transmission, the controls are simplified to the point where we can manage them without thinking.

But manufacturers have eliminated one set of distractions only to replace them with another. Like a military plane, the instrument panel of an up-market saloon is covered with switches and indicators, except they are dedicated to information and entertainment. There are controls for the satellite navigation system, the climate control system, radio, CD player, and hands-free cell phone. To activate a function while on the move requires the driver (a) to remember which control is which, and (b) to find the control without taking his or her eyes off the road. The process may last for several seconds, potentially distracting attention away from a hazard long enough for a crash [7]. This is why gadgets in the car need to be designed in a radically different way from gadgets in the home. But specialists agree that even a clever design cannot redeem the cell phone. Using a phone (hands-free or otherwise) is not compatible with the driving processs because having one’s eyes on the windscreen is not the same as having one’s attention on the road [9].

### Eco-driving

Some governments in Europe are promoting the notion of ‘eco-driving’ at national level on the grounds that if only people were aware of it, small adjustments to the way they drive would save a significant amount of fuel. It has been suggested that by driving in the highest possible gear and accelerating less hard than they might otherwise do, drivers can cut their fuel costs by 20% [3].

The chief saving comes from cruising at a modest speed using a fraction of the available power. In practice, cars are not optimised for cruising in this way. When the design parameters are fixed during the development of a new model, the engine capacity is determined by acceleration performance, not cruising performance. The engine must be large enough to cater for the relatively small number of occasions when maximum power is needed – overtaking, climbing a steep hill, or accelerating away from the traffic lights. Consequently it has twice as many cylinders as it needs to keep things ticking along. And since under motorway conditions, fuel consumption per unit distance travelled increases roughly with the square of the vehicle speed [4], a small speed reduction yields a large benefit both in terms of fuel costs and carbon emissions. On a level motorway where the traffic is flowing freely, the cruising speed at which the fuel consumption is minimised is usually in the region of 50 mph (80 km/h).

One can also save fuel in urban traffic by avoiding unnecessary braking and acceleration. Imagine you are driving along an arterial road and see a traffic signal in the distance change to red. You judge that it will be some time before it switches back to green. What would you do? The eco-driver eases off the throttle straight away, coasting slowly up to the signal so that the car is still moving when the lights change. This will annoy everyone behind, but a driver who jerks to a halt has failed to anticipate events, and the energy that the fuel has put into the vehicle mass is now dissipated into the surrounding atmosphere. Brakes, in fact, are a kind of litmus test. Other things being equal, the more you need to use them, the more energy you are burning. Aggressive driving merely heats the neighbourhood.

To show what can be achieved, in 2010 a team assembled by the Sunday Times newspaper and led by Gavin Conway broke the world record for distance travelled in a production car on one tank of fuel. The car was a Volkswagen Passat BlueMotion 1.6 TDI diesel with a 70 litre fuel tank. It featured low-resistance tyres, start-stop engine cut-out, low ride height, regenerative braking, higher gear ratios, modified aerodynamic styling with the drag coefficient reduced from 0.29 to 0.27, and no spare wheel. The drivers were careful not to accelerate rapidly, cruised at a little under 50 mph, and did not use the air conditioning [12]. They managed to drive 2457 km at an average fuel economy of 2.85 litres per 100 km - over 90 miles to the gallon.

Revised 23 February 2015