Resilience Scores for Decision Support Using Wearable Biosignal Data with Requirements on Fair and Transparent AI

Industry 5.0 brings human, AI, data, and robots together by supporting decision makers thus positively affecting the work balance between workers and machines. One key aspect for the daily decisionmaking on worker allocation in production is to consider the resilience of individual workers. Advanced manufacturing companies intend to foster wellbeing while avoiding illness, absenteeism and loss of resources with economic impact. We present workinprogress on introducing human resilience as a central human factors construct for optimizing longterm effects of humanmachine interaction on worker's health. For this purpose, components for computing AI-based physiological strain as well as resilience scores were developed. We also project towards adequate optimization methodology to adjust worker allocation based on resilience in time. Motivated humancentered production highly depends on the acceptance based on fairness and transparency, specifically in the context of the novel AI technologies.