Students racing over curving roads in a video-game simulation are helping scientists at the Lighting Research Centre (LRC) at Rensselaer Polytechnic Institute answer a serious question: Do current lighting standards mislead us about the efficiency of roadway, parking lot, and security lighting?

Lighting efficiency is widely calculated as lumens per watt of energy, explains Mark Rea, LRC director. The lumen, a measurement defined in the 1920s, is based on the response in bright light of the fovea, the part of the retina that contains cone photoreceptors and is responsible for central, high-acuity vision.

But another part of the retina containing rod-shaped photoreceptors is important in low light conditions and for peripheral vision. As a result, lumens per watt is an accurate efficiency measure for tasks done in bright light, but is less accurate for tasks such as night driving, which require good peripheral vision in low light.

In fact, using lumens per watt for night-time lighting can be very misleading. High-pressure sodium lights, for example, give more lumens per watt than metal halide lamps, but in low-light conditions, metal halide lights produce better reaction times to objects seen by peripheral vision. Metal halide lamps also produce far better colour-naming accuracy, a key factor in security lighting.

Optimising night lighting is a complex question that must also consider such issues as vision during transitional times such as twilight and response to the glare of oncoming headlights, Rea says.

In research sponsored jointly by General Electric, OSRAM Sylvania, Philips, and the Department of Energy, the LRC is conducting experiments with the driving simulator to measure the reactions of participants under varying lighting conditions.

The information produced is expected to help industry produce new lighting systems that are more efficient because they take into account the complex responses of the human eye.