The relentless drive for efficiency in Industry 4.0, coupled with persistent labor market pressures and supply chain volatilities, has intensified the demands on manufacturing workers. Traditional shift scheduling and task allocation methods often fail to account for the dynamic cognitive and emotional states of employees, resulting in increased risks of burnout, fatigue, and diminished overall well-being. Smart manufacturing environments, with their rich data streams and potential for AI-driven optimization, offer an opportunity to address these challenges through more human-centric workforce management. This paper introduces and evaluates an AI-driven framework for cognitive-aware shift scheduling and task allocation designed to mitigate worker fatigue and enhance emotional well-being in smart manufacturing settings. The primary objective is to investigate the impact of this AI-optimized approach compared to traditional scheduling methods on measurable cognitive load, subjective fatigue, and indicators of emotional well-being. The research addresses a significant gap: while AI has been applied to workforce optimization, few frameworks dynamically integrate real-time or predictive cognitive load and fatigue markers into the scheduling and task allocation process, and empirical evidence of their impact on worker well-being is limited. A mixed-methods experimental study was conducted in a simulated electronics assembly plant. Participants (N = 80 manufacturing workers) were assigned to either a control group (traditional, fixed-rotation scheduling) or an experimental group (AI-optimized, cognitive-aware scheduling) for a period of 4 weeks. The AI framework utilized a hybrid approach combining constraint-satisfaction optimization for schedule generation and a reinforcement learning agent for dynamic task allocation, informed by predictive models of cognitive load (derived from historical performance and task complexity) and real-time fatigue indicators (simulated from wearable sensor inputs like heart rate variability (HRV) and electrodermal activity (EDA), and subjective reports). Data were collected on objective cognitive load (via a validated secondary task reaction time (STRT) paradigm and simulated EEG-derived workload indices), subjective fatigue (using the Karolinska Sleepiness Scale and Stanford Fatigue Scale), emotional well-being (as measured by the WHO-5 Well-Being Index), and production output.

The AI-optimized, cognitive-aware scheduling and task allocation framework demonstrated significant improvements in worker well-being and manageable impacts on productivity. Workers in the experimental group exhibited a statistically significant reduction in average daily cognitive load. They reported lower levels of fatigue and higher scores on emotional well-being compared to the control group. Specifically, the AI framework led to proactive adjustments in task assignments and micro-break suggestions, which correlated with more stable performance patterns and fewer instances of extreme fatigue. AI-optimized scheduling reduced measured cognitive load (STRT latency) by 15.3% and emotional fatigue (Stanford Fatigue Scale composite) by 22.8% over the study period while maintaining 97% of baseline productivity levels.

This research provides strong evidence that AI-driven, cognitive-aware shift scheduling and task allocation can be a powerful tool for enhancing the mental and emotional well-being of manufacturing employees without compromising operational efficiency unduly. The findings support the development and deployment of human-centric AI systems in smart manufacturing, moving beyond purely output-driven optimization to consider the cognitive and emotional sustainability of the workforce. This has critical implications for designing healthier, more resilient, and more productive manufacturing environments in the industry 4.0 era.