Using Probabilistic Reasoning to Develop Automatically Adapting Assistive Technology Systems
Richard C. Simpson, Simon P. Levine, Heidi Horstmann Koester · 1996
Automatic adaptation has the potential to enhance the performance of a wide variety of assistive technologies. This paper describes two adaptive assistive technologies: (1) a "smart wheelchair " that selects its operating mode based on current task requirements and (2) a row-colunm scanning keyboard that changes its scanning rate to conform to the capabilities of the user. The reasoning method underlying each system’s adaptation decisions, a probabilistic reasoning construct known as Bayesian networks, is also discussed.