Fatigue-Aware Learning to Defer via Constrained Optimisation
This work addresses the problem of human-AI collaboration in decision-making for domains where human fatigue affects performance, offering an incremental improvement by incorporating fatigue modeling into existing L2D frameworks.
The paper tackles the problem of learning to defer (L2D) by addressing the unrealistic assumption of static human performance, proposing FALCON to model fatigue-induced degradation and optimize accuracy under human-AI cooperation budgets. Experiments show FALCON consistently outperforms state-of-the-art L2D methods across coverage levels and generalizes zero-shot to unseen experts with different fatigue patterns.
Learning to defer (L2D) enables human-AI cooperation by deciding when an AI system should act autonomously or defer to a human expert. Existing L2D methods, however, assume static human performance, contradicting well-established findings on fatigue-induced degradation. We propose Fatigue-Aware Learning to Defer via Constrained Optimisation (FALCON), which explicitly models workload-varying human performance using psychologically grounded fatigue curves. FALCON formulates L2D as a Constrained Markov Decision Process (CMDP) whose state includes both task features and cumulative human workload, and optimises accuracy under human-AI cooperation budgets via PPO-Lagrangian training. We further introduce FA-L2D, a benchmark that systematically varies fatigue dynamics from near-static to rapidly degrading regimes. Experiments across multiple datasets show that FALCON consistently outperforms state-of-the-art L2D methods across coverage levels, generalises zero-shot to unseen experts with different fatigue patterns, and demonstrates the advantage of adaptive human-AI collaboration over AI-only or human-only decision-making when coverage lies strictly between 0 and 1.