The influence of individual characteristics on task execution time prediction under variable time constraints: A within-subject study

In complex industrial environments, modelling human performance is fundamental to ensuring safety and efficiency, particularly in extraordinary maintenance activities. Consistent with these challenges, this study examines the predictive value of nine individual characteristics in determining the completion time of a cognitive-motor task. Twenty-two operators, divided between experts and non-experts, participated in a within-subjects laboratory experiment, performing a maintenance simulation on a vacuum-filling machine under two conditions: without time constraints and with a seven-minute time limit. The data collected, relating to time pressure and performance in carrying out the activities, were analyzed using linear regression, Lasso Regression, and Random Forest models. The analysis showed that traits such as manual dexterity and decision-making ability are significant predictors of performance, although their influence varies under time pressure. In particular, the inclusion of contextual variables improved the interpretability and accuracy of the models. These results underscore the importance of developing human-centred predictive models that integrate individual characteristics with environmental factors, thereby contributing to the design of adaptive systems in high-risk industrial contexts. The study also emphasizes the importance of incorporating contextual variables into predictive models. It offers insights for the development of adaptive workforce management and training systems in line with the principles of personalization and resilience promoted by Industry 5.0.