Sustainable Work Allocation Under the Three-Compartment Muscle Fatigue Model: Invariance, Closed-Form Rest Allowances, and a Certified Shield for Human–Robot Collaboration
Year: 2026
Type: article
Abstract: Allocation schemes for human–robot collaboration track operator fatigue, but the underlying variable is rarely analysed, so none can say what it guarantees about the operators being scheduled. Here, the three-compartment muscle fatigue model with rest recovery (3CC-r) is treated as a switched dynamical system and the guarantees derived from it are presented. The well-posedness and forward invariance of the probability simplex are established; the equilibrium under constant load and the largest sustainable load follow in closed form. Every unsaturated work–rest duty cycle converges to a unique, globally exponentially attracting periodic regime with per-cycle contraction, exp[−R(Tw+rTr)], from which closed-form minimum rest allowances and maximum work bouts follow. A sharp step-size criterion identifies when explicit Euler preserves the simplex; the common 100 ms control period falls just outside it for the rates used here. These results yield a fatigue shield: a closed-form admissibility test, appendable to any policy, that provably keeps every operator’s fatigued fraction below a prescribed threshold under the reduced fatigue dynamics. Across 360 simulated shifts in a three-operator, two-robot cell, the full 3CC-r model was observed to remain below threshold in every shielded shift under three different allocators, at a throughput cost of at most 2.5%; unshielded, fatigue reached the saturation value of 0.915. The analytical guarantee presumes calibrated rates; inter-individual variation calls for conservative ones. The code and the reproducible cell are released.
Source: Mathematics
DOI: 10.3390/math14224278
Author: Claudio Urrea
Institution: Universidad de Santiago de Chile
Language: English
Topic: Ergonomics and Human Factors
Subfield: Human Factors and Ergonomics
Field: Social Sciences
Domain: Social Sciences
Open Access status: gold
Estimated APC paid: $3,137
OpenAlex ID: W7237208149
