Transparency Enhancement in Teleoperation: An Improved Model-Free Predictor for Varying Network Delay in Telerobotic Application

Abstract

This work proposes a useful modification to an existing model-free predictor for handling network delays in teleoperation. This modification consists of an adaptively varying parameter that depends on delay at the current instant and the latest prediction error terms, which was not in use in the existing method. The conception of the proposed modification was obtained through theoretical analysis and later implemented for a mobile robotic platform. It was observed that the modified predictor adapted well to the varying delay and improved the system transparency by a maximum of 86% and by 15% in average. Additionally, this work also proposes a method to compensate for missing data using the same model-free predictor, showing a maximum transparency improvement by 12%.