Adaptive Task Support Based on Dynamic Human State Estimation

Adaptive systems that provide task support to the human operator when needed can be a solution to the problems of traditional automation in complex dynamic systems. One way of initiating support is automatically: let the technical system assess operator functional state based on continuous monitored physiological, performance, and environmental factors and let it decide whether support is needed. For this concept to work in practice several human-related problems need to be solved. This is the focus of the REFLECT project which aims at developing adaptive systems that really support the human operator. Some of the accumulated knowledge gained in this and preceding projects are addressed in this paper. In addition, a design cycle for adaptive systems is presented that provides a functional overview of the design issues to be solved.