Temporal dynamics of movement patterns : Expertise and behavioural variability – DYnaMov

This project proposes a multi-disciplinary approach to the analysis of human movement to explore behavioural adaptability as a determinant property of expertise. It combines knowledge and methodologies in the human movement sciences, computer sciences and applied mathematics. The primary aim is to investigate the functional role of movement variability in natural contexts where uncertainty and temporal pressure are high, requiring that individuals constantly adapt their behaviours in order to respond to existing dynamical and interactive constraints. Studying the functional role of movement variability involves assessing how adaptive is human behaviour by analysing the balance between movement pattern stability (i.e., persistent behaviour) and flexibility (i.e., variable behaviour) relative to a performance context. Specifically, our project explores how experienced and inexperienced individuals in work and sport, such as firemen and athletes, adapt their motor behaviours in various performance contexts, i.e., when the environmental properties are stable or in unstable or transitional regions where movement patterns co-exist (technically known as a metastable region). The experimental aims include exploring how environmental properties (performance in stable vs. meta-stable region) influence behavioural adaptability, when inexperienced firemen and athletes are trained either in a stable environmental constraint or in meta-stable performance regions.

Society demands the rapid and effective minimization and management of high risk performance situations. Hence, training in meta-stable performance regions is particularly relevant in order to facilitate those tasked with responding to high levels of uncertainty and temporal pressure. Therefore, the findings of the project could provide useful knowledge to design training programmes requiring continuous motor adaptation, including exploration and escape strategies (i.e., a fall back to safer or more reliable movement patterns).
To permit the analysis of motor behaviour adaptability, the objective for computer sciences and applied mathematics relate to the development of movement pattern recognition software in order to track human behaviour in space-time dynamics by identifying “attractors” (i.e. stable movement patterns) and transitions between attractors. The method chosen for this project is the use of clustering analysis to study inter-individual variability and to detect individual profiles, and Hidden Markov Model HMM to analyse the temporal dynamics of movement patterns. An additional focus will be on critical points of the temporal dynamics of movement patterns by analysing and modelling global and local variability of the time-series. This cooperation between computer sciences and applied mathematics and human movement sciences would enable the development of a multi-disciplinary team of research, across regional, national and international levels (i.e. Keith Davids, with QUT in Australia, is a world class researcher renowned for developing a novel and fruitful framework to understand the coordination of actions with events in dynamic performance environments).

In summary, this research proposal describes a developmental experimental approach, whereby the main technical challenge is to develop a measurement system capable of capturing human behaviour in natural performance contexts using embedded sensors. Concurrently a method will developed for the analysis of human movement data using a small number of parameters. The project attempts to use inertial measurement units (IMU) which associate 3D accelerometer, 3D gyroscope and a 3D magnetometer. This project proposes the use of MotionPod3 by MOVEA (Movea©, Grenoble, France, movea.com) who will contribute as a partner in the project (i.e. not in a service delivery role).