Cognitive assistance for action selection: Challenges and approaches

Abstract

Cognitive assistance systems aim at compensating shortcomings of natural cognition concerning specific activities. Notable progress has been made regarding data acquisition, analysis, and the exploration of technical means for supporting human action selection and execution. The related challenges and potential solutions can be associated to four largely independent questions: What actions should be executed, when this must or should be done, whether assistance is needed for a specific action, and if so, how the action should be supported. A broad range of technological and methodical approaches can be taken for tackling each of these issues, including recent advances and new challenges in the automatized analysis of task-related mental representation structures.