Modeling Human Sensory-Motor Action for Vehicle Braking Based on Fuzzy Inferences

The mental modeling in terms of sensory-motor changes is one of the most challenging task in artificial inteligence and definitely is an important milestone for engineering and neurosciences. This paper starts from a reverse engineering perspective comparing functionality between real system - the brain and target system - the simulation model. At the conceptual level, we particularly analyse the three stages involved in mental functional mode of human action: sensory, perceptual, and representational (SPR). We proposed some structural models for basic functional modules and for the entire sensory-motor chain. The basic evidences that motivate us to consider fuzzy systems as suitable models for mental mechanisms in terms of sensory-motor changes are discussed. In this research we developed some minimal fuzzy inference systems simulating sensory, perception and representation blocks with specific fuzzified inputs. The three blocks connected into so called SPR structure is simulated in a certain scenario of human action and the results are analyzed from the perspective of human consistency. The proposed model is designed as a general case to study the action mode based on visual stimuli with common task to avoid the collision with an obstacle by controlling braking in car driving. The results show good agreement between simulated signals and human reactions in context.