Title: Real-Time-Human-Performance-Monitoring
Approval Date:  March 21


In a collaborative context with experts from Crossing (AI, Human Factors, Signal Processing, Embedded Systems) the objective is to develop a state model (stress, fatigue, attention, performance) of a human operator with a real-time sensor fusion architecture using physiological, behavioral and environmental data. After an in-depth study of available sensors (on-body or in the environement) and associated techniques from the state-of-the-art, we will establish a robust experimental design and adaptable statistical model to facilitate measurement of the key relevant metrics. The development of the state model may benefit from machine learning to capture and understand individual and hybrid team performance from multi-modal data sources.The ultimate aim is get a model to used later to adapt, according to a team performance criterion, the distribution of tasks (load, complexity) between an operator and autonomous agent (AI).

Related roadmap thrusts and axis

– Thrust 1: New Models to Understand and Anticipate Human Behaviour

  • Axis 1.1, 1.2

– Thrust 2: New Algorithms for Energy-Efficient and Human Based AI

  • Axis 2.3

Project Members

Name Organisation  Role
Dr. Nathan Beu CNRS  Main investigator
Prof. Anna Ma-Wyatt The University of Adelaide  HF, psychology
Prof. Cedric Buche CNRS, ENIB
 AI, Human/Machine
As/Prof. Philippe Rauffet CNRS, UBS
 HF, Human Performance monitoring.
Dr. Jean-Philippe Diguet CNRS  Real-Time, Embedded systems, Sensor fusion
Prof. Siobhan Banks University of South Australia  HF, Fatigue.
 non exhaustive list ….  


Project Funding 

  • AID

Start / Duration 

  •  Nov. 2021, 18 Months years