The task execution in the multi-user telepresence systems often require two or more teleoperated robots to handle with the same object. The interaction between them lacks of true transparency especially in the haptical channel, which leads to several objectionable issues like high failure ratio and unnecessary operator load.

These negatives can be compensated and the cooperation quality can be improved if the autonomy of the teleoperated robots increases. It is to be done not by training the robot to perform some predefined tasks, but to generalizing the tasks and thus the right balance between teleaction and autonomous handling to be found.