Taking a shine to safety

Car makers can now specify visibility standards for in-car display screens from their suppliers. This should improve road safety as well as customer satisfaction — twin results that every car manufacturer craves.

A screen that cannot be read quickly by a driver can, at best, lead to frustration. At worst, he or she can be so distracted that their concentration is broken and safety is compromised.

Until now car makers have not been able to specify measurable scientific criteria that the screens must achieve. But researchers at the National Physical Laboratory (NPL) at Teddington, Middlesex, have built a dedicated facility and devised a rigorous methodology to solve the problem.

The project began a year ago when NPL's photonics group, led by Martin Wicks, received a request for help from a car maker.

'There had been a lot of work on display visibility for aerospace, but nothing applicable to the type of displays in cars,' said Wicks. 'There's increasing use of LCDs in cars for satellite navigation units and to replace traditional instruments such as speedometers.'

Fortunately a large enough enclosed space — the size of six domestic garages, and twice the height — became available at NPL. The challenge was first to design a lighting set-up that would be a good simulation of natural daylight, and then to devise a methodology so that subjective assessments could be compared and correlated with scientific measurements.

'We need to simulate the environment of a bright sunny day, when there is about 100,000 lux of light,' said Wicks. 'And it's not just direct sunlight. There's a lot of reflected light.'

The team opted for 4kW arc lamps operating with the daylight spectrum, and four fabric screens on metal frames. 'The screens reflect 95 per cent of the light,' said Wicks. A special saucer-shaped light, simulating an artificial sun, was also specified.

But the team soon decided that even more lighting should be installed. 'We didn't want the lamps to be too close to the subjects because that could cause discomfort and distract them from the task in hand,' said Wicks. 'That's when we realised we actually had to increase the capability to achieve the desired lighting levels.'

In the end they opted for five 4kW arc lamps. A consequence of this decision was a need to increase the cooling of the facility. 'It meant we had to upgrade the air conditioning because we were dumping a total of 20kW into the lab,' said Wicks.

With the lights set up, a car with a display screen was driven into the lab. A telespectroradiometer was used to measure the intensity of light and its distribution across the spectrum. 'It looks at both the light emitted from the display and reflected by the screen,' said Wicks.

The team determined that, for the first cars to be tested in the new lab, the most difficult condition for reading the display would be when a car was heading straight towards a low sun. So the condition was simulated with the arc lamps and reflector screens.

Then it was time to bring in the human volunteers. Subjects were asked to adjust the brightness of the display to a point where it was just readable, to define the visibility under the worst lighting condition.

They were also asked to make a second adjustment to the lighting level of the display so that the screen was easy to read. Software captured the settings that each person chose.

The final step in the methodology was to correlate subjects' assessments of the display's visibility with measurements from the telespectroradiometer. 'The readability of a display is the effective contrast between the brightness of the display and the amount of light reflected off it,' said Wicks. For the first time, his team has formalised a method and built a lab that can gauge this repeatedly and scientifically.

'Now car makers can tell their display suppliers the criteria that the products must meet,' said Wicks. 'There's the safety benefit and also the customer satisfaction issue. People pay a lot for their new vehicles and they want them to be good.' A hard-to-read display would lose points as soon as it is driven out of the showroom.

The NPL lab is also suitable for measuring the readability of other displays, and Wicks expects to win business from those who specify instrumentation for systems such as train cabs, boats, signal boxes, information kiosks and large outdoor screens.