The goal of this research is to realize robust teleoperation with multiple telerobots under changing communication topologies. In haptic teleoperation with multiple operators and multiple robots the forces as well as the positions should be transmitted between all the robots, which introduces difficulties in maintaining the stability and the efficiency of the system under communication failures or time delays on certain paths. In our research, we relax the constraints on the control system by approaching the multilateral teleoperation problem by using information graphs and consensus algorithms. We first show how decentralized multilateral teleoperation can be realized by using consensus filters. The haptic consensus algorithm which employs consensus filters is shown to be effective even on dynamically changing network topologies, as long as the network structure is connected and balanced. Furthermore we show that using haptic consensus algorithm it is possible to realize a bilateral teleoperation system for cooperative load carrying. Experiment results show that the haptic consensus algorithm enables multilateral teleoperation as well as bilateral multirobot cooperative teleoperation even on switching network topologies where the network topology is changing every control sampling time, which in our case is 100 microseconds. This means that the control system is very flexible, and tolerant against communication failures.