Using biomathematical models to assess rosters for fatigue
Our research in this area focuses on enabling the rail industry to reduce the risk of designing fatiguing rosters. The results of our analysis of BMMs based on industry needs is available to organisations to inform the selection of tools to support fatigue management.
While biomathematical fatigue models can be very useful tools to support worker scheduling, they present a number of limitations that need to be borne in mind. For example, they produce estimates of fatigue based on assumptions of how people might sleep if they work different rosters, but in reality, people get different amounts of sleep at different times when they work shifts. The cumulative effects of hours of work are also not well represented in the models.
This means that the output of fatigue risk tools should only constitute one component of the fatigue risk assessment. Outputs should be interpreted with caution, as they do not provide an absolute measure of fatigue. They should not be used in isolation to make go/no-go decisions.
We have developed good practice guidance to help rail industry staff develop a better understanding of what biomathematical fatigue models are, what their limitations are and how they can be integrated within a wider fatigue risk management system. We recommend that individuals that use these models receive training on fatigue and fatigue management; that model users avoid simplistic interpretations of the numbers the models provide; and that any thresholds that are used are validated for the specific activities of the company. Ultimately, biomathematical models should be used to compare different work schedules, rather than conduct direct evaluations of a schedule.
We have used the outputs of this research to design the Planning and rostering section of this site.