Driver approach behaviour at seven railway level crossings in Ontario was recorded on video tape during 301 hours. The crossings varied according to protection (crossbucks only, flashing lights and bells and gates) and urban versus rural location. Hourly traffic flow (in both directions combined) ranged from a maximum of approximately 900 to a minimum of about 2 vehicles. The records were analyzed qualitatively for the occurence of critical incidents, defined as events that bring an accident appreciable closer to its potential realization. Most of the critical incidents took the form of road users either inadvertently of deliberately entering the crossing after signal onset. Critical incidents were common events; they were seen in about tho-third of all periods in which warning signals were activated. Unduly long warningh times, long occupancy times of the crossings by some trains and false alarms were interpreted to stimulate the tendency of road users to deliberately violate the signals. Inadvertent violation of the signals were was explained as a result of the physical impossibility to come to a safe and comfortable stop in front of a crossing which inevitable applies under certain conditions. Driver approach speeds of about 3200 lead vehicles were analyzed quantitatively from the video records and showed high coefficients of speed variation between drivers as well as within drivers as they moved closer to the crossings. During a rail strike drivers approached crossings at a higher speed than they did under normal conditions. (A)
Abstract