Teleoperation remotely moving?

Citation

Pegman, G. (2003), "Teleoperation remotely moving?", Industrial Robot, Vol. 30 No. 6. https://doi.org/10.1108/ir.2003.04930faa.002

Teleoperation remotely moving?

Teleoperation – remotely moving?

Geoff Pegman

Geoff Pegman is the Managing Director of RURobots, PO Box 248, Manchester M28 1WF, UK. Tel: 0161 799 3898; Fax: 0161 799 8328; E-mail: geoff.pegman@rurobots.co.uk; Web site: www.rurobots.co.uk

Keywords: Teleoperation, Haptics, Master-Slave control

Teleoperation has the longest heritage of any industrial robot control system. Ever since 1947, when the Argonne labs in the USA developed the first master-slave systems, teleoperation of manipulators has been firmly on the industrial landscape. The provenance of vehicle teleoperation is somewhat murkier, but dates back to at least the early 1940s.

The abilities of the combined human-machine system have seen teleoperation systems become firmly entrenched in the hazardous environment areas such as the nuclear industry, subsea and military applications. Given the time that these systems have been around, and the importance of the areas in which they are used, it would be reasonable to assume that the human interface must be nearing a state of perfection. If this is true, then a review of systems in use would suggest that the perfect interface is a positional mini-master or rate joystick and a small CCTV view. The author not only doubts that this is the best interface for controlling current robots, but also believes that the acceptance of these interfaces is restricting the development of robotics into new applications.

In the last 30 years, robot systems have undergone tremendous development. Drive systems have become more compact and efficient, structural properties have been improved and control system capabilities have been transformed. And yet the main way of controlling these robots has scarcely evolved at all during the last 20 years. Indeed, in moving from direct wire or tape driven master-slave systems to electronic control systems has largely meant giving up any force feedback and a total reliance on only the visual sense for any feedback. This is not to say that developments have not taken place both in the research labs and in products offered for sale. It is just that, in the main, these developments have either not reached the market or not been taken up by the users. Even relatively simple enhancements with apparent benefits, such as “virtual cameras” using real-time 3D simulation displays, have had limited take up amongst the traditional user community.

Only in new areas such as medical robotics are the ways in which users interact with robots being really examined and developed. In such areas, haptic feedback assumes an importance almost as great as visual feedback. As importantly, the ergonomics of the interface are investigated and optimised to provide an intuitive and precise control of the robot. If there are benefits to be gained in the medical arena, would not the nuclear, subsea, and military areas benefit from being able to take on more dextrous tasks remotely rather than resorting to the suited operator, diver or well-padded Explosive Ordnance Disposal Officer to carry out such tasks?

It is not only the physical nature of the interface that is changing little. Despite much research work in shared and traded control, most teleoperation systems still utilise direct teleoperation where the operator is severely tested to carry out parts of the task that would be trivial to carry out with the appropriate automation. Supervisory control is nowhere in sight.

Eventually teleoperation systems could evolve into two classes of personal robot. First is the voice command robot, where the operator issues simple, spoken language supervisory commands. Second is the alternate-ego telepresence robot that will allow the user to explore distant places and experiences. Whether these will be developed through pushing the barriers of current teleoperation systems into new applications areas, demanding more advanced and more effective user interfaces, remains to be seen. It could be that these ultimate teleoperation systems arise instead from adaptations of autonomous robots and virtual reality systems. If that is so, we will not only have to wait longer for these systems to emerge, but many useful robot teleoperation niches will remain unfilled. At present, the best prospect for the widespread adoption of more effective teleoperation systems is for the traditional users to pay heed to what is happening in the emerging application areas.