Telemanipulation enables a user to operate in a remote environment by using a robotic system. However, challenges such as enabling a stable and transparent system under time delays must be overcome to achieve that goal.
Model-mediated Teleoperation (MMT) seeks to address this problem by avoiding traditional closed-loop control; instead, the MMT approach uses data from the teleoperated environment to construct and update a competent model of its environment. While the MMT literature provides solutions for an environment with a single type of dynamics, the dynamics in teleoperation in complex environments are expected to switch depending on the type of interaction.
This work explores how an MMT system can support an environment with changeable dynamics. In this study, two designs are created and experimentally tested to identify and switch dynamics in an environment with changeable dynamics. To achieve that, an environment with changeable dynamics is modelled, where the type of information available on the environment is the main source to inform model switching for an MMT system.
The effectiveness of the two designs, along with the modelling accuracy of the proposed environment, are assessed, showing the conclusion and future lines of research for the proposed MMT design.