An Evaluation and Analysis of Master Slave Controllers for Dexterous Manipulators under Time Delay and the Effects on Perceived Mental Workload and Task Performance

Robotic manipulators that use joystick interfaces for telemanipulation take time and effort to master and lack an intuitive basis for human robot interaction. These factors can hamper operator performance, increase cognitive workload, and limit overall user effectiveness on highly dexterous tasks. Master-slave interfaces (MSIs) used in the fields of animatronics, telesurgery, and hazardous materials handling are intuitive and improve operator performance, but are limited because they typically operate in a controlled environment and lack robustness. When MS controllers use force-feedback errors can accumulate and create instabilities due to the synchronous bilateral communication between the remote and local systems. Instability can be mitigated and errors reduced in these systems by separating the bilateral communication into a dual-unilateral (DU) asynchronous communication which alters the way the interface and remote system communicate. This research focused on comparing and analyzing current MSI controllers with a new DU control method that reduces or eliminates MSI issues such as latency, error, and information loss while providing the user with accurate force-feedback capability and transparency during operation. It used a simulation-based approach to determine which selected MSI controllers provide users with robust control of dexterous manipulators under time delay. Users performed a variety of dexterous force-feedback tasks with different controllers and the study measured objective task performance and completion time success along with cognitive workload related to user perceived task interaction and mental demands through subjective ratings on the NASA (National Aeronautics and Space Administration) Task Load Index (TLX). The research findings helped to develop a comprehensive and scalable testbed for MSI controllers with the focus of providing operators with a robust and intuitive controller for low-level control of dexterous manipulators and establish future guidelines for conducting these tests and integrating controllers and interfaces into a virtual simulation.