Movement variability and digital human models: development of a demonstrator taking the effects of muscular fatigue into account.

Movement variability is an essential characteristic of human movement: the same motor task is never performed in exactly the same way twice. This variability is related to the numerous redundancies in the human locomotor system, and depends on several factors, some of them linked to the individual and to the task to be performed. For example, muscle fatigue can cause postures and movements performed to complete a task to change during a work day. Despite its prevalence, movement variability is almost completely ignored in workstation design, where designers generally aim to define a unique standard procedure optimising the production cycle.
This presentation starts with a review of the current state of knowledge on movement variability, how this is related to the different theories of motor control and the factors influencing it. Movement variability occurring in occupational activities will be highlighted and the importance of taking it into account from the stage of workstation design will be discussed. Indeed, neglecting this movement variability can lead to poor planning for the future activity performed by operators, thus leading to incomplete assessment of biomechanical risk factors. Being able to simulate the future operator’s movement variability from the stages of workstation design will therefore contribute to improving the prevention of occupational risks.
Following this review, a demonstrator is described. It is intended to simulate, as an example of feasibility, movement variability induced by muscle fatigue from the design stages of a workstation. This demonstrator is based on the XDE virtual human simulation framework which has previously been used to analyse movements, torques and efforts as part of ergonomics analyses. Its multi-objective dynamic controller can handle several simultaneous tasks (balance, contacts, manipulation tasks or even locomotion) while respecting the laws of physics. This virtual human is enriched with a dynamic model of muscle fatigue, which is known to contribute to motor variability. The model chosen for fatigue and recovery phenomena is based on a three-compartment model. It calculates a joint fatigue index which takes the dynamic state of the model (angles, speeds and joint acceleration) into consideration, as well as an instantaneous maximum torques approximated by the OpenSim musculoskeletal simulation framework.
This demonstrator is intended to show that it is feasible to develop tools to assist the design of workstations which take movement variability into account for improved ergonomics at the workstation.