Abstract: Embodiments pertain to a computer-implemented method of predicting muscle fatigue in a subject by: (1) receiving data related to muscle fatigue in a muscle of the subject; (2) feeding the data into an algorithm to predict the muscle fatigue; and (3) outputting the predicted muscle fatigue. The algorithm my include: (a) an active compartment (M_A) model representing individual motor units (MUs) generating force at full capacity, (b) a resting compartment (M_R) model representing inactive MUs capable of rapid activation into force production, (c) a centrally fatigued compartment (M_FC) model representing fatigued MUs due to a central mechanism dominant at zero or near-zero joint velocities, with rapid recovery, and (d) a peripherally fatigued compartment (M_FP) model representing fatigued MUs due to a peripheral mechanism dominant at higher velocities, with slow recovery. Further embodiments pertain to a computing device for predicting muscle fatigue in a subject in accordance with such methods.