From the end user point of view, the human-robot interface is often the most important subsystem of an intelligent machine. In multi-task service robots its role is emphasized because of the versatility and increased autonomy and complexity of the system. Because of the limited intelligence capabilities of the robot, it has to ask for help from the operator, especially in tasks where complicated perceptive and cognitive skills are needed. The robot has to include the operator as a part of the task control loop. This is the challenge in human-robot interaction. Human effort is needed in complicated and cooperative tasks, but it should not consume too much operator time or be exhausting for the operator. Additionally, the parameters of this optimisation are totally task-/application-dependent.

In the research and development of human-machine interaction it is important to utilize all the features which modern information and multimedia technology provides. There is a strong tendency to change the role of the operator of the machine from steering or passive monitoring to interactive communication, where the operator transmits his will to the machine using mainly semantic information and means natural to humans.

At ATL research in this direction has been already carried out over a period of several years under the heading “cognitive human-robot interface”, where a symbolic level communication interface has been created. The underlying idea is to utilize the concept of “spatial situation awareness”, which puts both the human and machine entities into the same virtual presence. With the help of this situation awareness model, all position-related objects can be perceived and understood symbolically by both entities, which in fact makes possible the use of symbolic-level commands in machine control. The actual need to control the machine - which can vary from teleoperation to highly autonomous operations - determines the details and technical aspects of the human-machine interface. Hitherto only the first steps have been taken in this field and there is much to do before practical results are available.

Both partners carry out this RP in cooperation. IHA has long experience in machine teleoperation and model augmented teleoperation. ATL has concentrated on telepresence and cognitive human-robot interfaces.