Intent inference in shared-control teleoperation system in consideration of user behavior

Abstract In shared-control teleoperation, rather than directly executing a user’s input, robot system assists the user via part of autonomy to reduce workload and improve efficiency. Effective assistance is challenging task as it requires correctly inferring intent, including predicting goal from all possible candidates well preferred movement in next step. this paper, we present probabilistic formulation for intent by taking consideration behavior. our approach, behavior learned demonstrations, which then incorporated prediction path planning. Using maximum entropy principle, two methods are tailored according similarity metrics between short-term movements We have validated proposed approaches with study—examining performance approaching tasks multiple goals scenario. The results show that perform able respond quickly dynamic changing goals. Comparison analysis shows outperform existing especially scenarios ambiguity.