Abstract
Driver behaviour has not been considered a high pay-off area for investment of safety research funds because of the difficulty in changing some of the obvious weaknesses. Nevertheless it will remain on area of interest, if only for its challenge to behavioural sciences. Modeling of driver behaviour is promising to become more fruitful as systematic analysis and criterion development are forthcoming. By controlling the relative attractiveness or resistance of alternatives, the road environment can help shape good driving behaviour. Artificial changes in resistance such as controlled visibility or use visual illusions, coupled with classification and indentification of drivers according to driving characteristics, may help in this goal.