Establishing optimal illuminance for pedestrian reassurance using segmented regression

The optimal illuminance from road lighting for pedestrian reassurance after dark is that beyond which further increase in illuminance has no significant effect on reassurance. Previous studies have not revealed a precise estimate of optimal illuminance. The current study investigates the use of segmented regression for defining optimal illuminance, applied to 25 662 reassurance evaluations by 380 test participants who recorded their reassurance assessments in 253 locations in three cities in Israel. Segmented regression led to models which better fit the empirical data than their unsegmented counterparts and offered precise estimates of optimal illuminance, these ranging from 8.9 lx to 26 lx, depending on location.