The next generation of fatigue prediction models: evaluating current trends in biomathematical modelling

Review of trends in biomathematical fatigue models (BMMs). Argues the next generation should move beyond population-average sleep/circadian models toward individualisation, real-time inputs and validation in the target operation.

Biomathematical models (BMMs) are parametric models that quantitatively predict fatigue and are routinely implemented in fatigue risk management systems in increasingly diverse workplaces. There have been consistent calls for an improved ‘next generation’ of BMMs that provide more accurate and targeted predictions of human fatigue. This article examines the core characteristics of next-generation advancements in BMMs, including tailoring with field data, individual-level parameter tuning and real-time fatigue prediction, extensions to account for additional factors that influence fatigue, and emerging nonparametric methodologies that may augment or provide alternatives to BMMs. Examination of past literature and quantitative examples suggests that there are notable challenges to advancing BMMs beyond their current applications. Adoption of multi-model frameworks, including quantitative joint modelling and machine-learning, was identified as crucial to next-generation models. We close with general recommendations for researchers, practitioners, and model developers, including focusing research efforts on understanding the cognitive dynamics underpinning fatigue-related vigilance decrements, applying emerging dynamic modelling methods to fatigue data from field settings, and improving the adoption of open scientific practices in fatigue research.

Verbatim from Semantic Scholar (opens in a new tab) · Abstract via Semantic Scholar; © the authors/publisher

No results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.