Dmitrij Polin

Robust Underestimation of Speed During Driving: A Field Study

Traffic reports consistently identify speeding as a substantial source of accidents. Adequate driving speeds require reliable speed estimation; however, there is still a lack of understanding how speed perception is biased during driving. Therefore, we ran three experiments measuring speed estimation under controlled driving and lighting conditions. In the first experiment, participants had to produce target speeds as drivers or had to judge driven speed as passengers. Measurements were performed at daylight and at night. In the second experiment, participants were required to produce target speeds at dusk, under rapidly changing lighting conditions. In the third experiment, we let two cars approach and pass each other. Drivers were instructed to produce target speeds as well as to judge the speed of the oncoming vehicle. Here measurements were performed at daylight and at night, with full or dipped headlights. We found that passengers underestimated driven speed by about 20% and drivers went over the instructed speed by roughly the same amount. Interestingly, the underestimation of speed extended to oncoming cars. All of these effects were independent of lighting conditions. The consistent underestimation of speed could lead to potentially fatal situations where drivers go faster than intended and judge oncoming traffic to approach slower than it actually is.

Reducing the stroboscopic effects of LED luminaires with pulse width modulation control

Pulse width modulation for dimming the light output of LEDs has become common. When pulse width modulation is used at low frequencies unwanted visual artefacts including flicker perception and stroboscopic effects may occur. These artefacts need to be avoided or at least reduced to a minimum in order to obtain high user acceptance. In this paper, an optimized phase-shifted pulse width modulation method is described, implemented and validated in a visual experiment. The method is intended to minimize the stroboscopic effect on a reference surface by first optimizing the LED units of a single LED luminaire and then co-optimizing several of these luminaires. The optimized pulse width modulation waveforms are then compared to standard pulse width modulation dimming methods. In the visual experiment, 13 subjects rated the extent of the stroboscopic effect of standard and optimized waveforms in a white painted experimental room. The results indicate that the optimized waveforms are indistinguishable from constant light.