System Analysis on Driver Monitoring System for Mainline Railway

Abstract

Driver alertness and attention are factors in nearly 50% of Signal Passed At Danger (SPAD) events that could lead to railway accidents.  My research aims were to carry out a system analysis on increasing the capability of the existing vigilance system for UK mainline passenger trains to include active detection and actuation based on driver’s alertness.  The current vigilance system has existed for a long time in UK railway history, and requires optimisation of its original capability.  The current vigilance operation can be tricked, isolated, or become a routine gesture.  There is a delay time of 60s with 5s action time for the driver to reset the ‘lack of activity’ trigger, which is far too long for trains operating at a typical line speed of 125mph.  This new capability is efficient in reducing the reaction time of detection on driver’s falling asleep by 97% (from 65s to 2s) for drivers who cannot be woken by audio alarm, and by 93% (from 30s to 2s) for drivers who can be woken up by a beeping sound.  Additional functionalities of the proposed design include detecting early signs of falling asleep, microsleep and eyes-off-road, both intentional and unintentionally.