Performance Evaluation of a Blended Shared Control Method in Human-Robot Teams

Presenter: Anas Abou Allaban

Research Category: Engineering and Technology
Student Type: null
PI: Anas Abou Allaban, Taskin Padir

When human-robot teams collaborate to complete a specific task, it is important to develop methods that enable them to share the decision-making process with one another. In this project, we investigate the implementation of a blended shared control (BSC) scheme for a mobile robot. Here, the human-robot team must work together to reach a certain location. BSC ensures that there is a human in the control loop monitoring the status of a robot, while the robot optimizes the human’s trajectory commands. We test the reliability and applicability of this approach by introducing environmental stimuli that inhibit the operator’s ability to control the robot properly. Here, environmental stimuli include either velocity drift or communication delay to the robot. The time to reach the goal and distance traveled are selected as performance metrics and are compared against a baseline trial with no external disturbances. We performed experiments with 12 human operators and it was concluded that BSC improved the operator’s control in the presence of velocity drift. However, no increase in performance was determined with communication delay. Overall, this study demonstrates that BSC can be an effective control scheme for human-robot teams, when the robot operates in a known environment.