Telerobotics

Submitted by aleleve on Fri, 03/18/2011 - 22:25

History

This activity started in 1996 :

This research activity drastically slowed down between 2001 and 2011 as I was working on E-Laboratories in ICTT and afterwards in LIESP laboratories, both dealing with e-learning research topics.

Since early 2011, this topic is restarting at AMPERE laboratory and in collaboration with LIRMM robotics team.

Introduction to teleoperation

There are situations when firms or laboratories have to resort to remote manipulation. Such cases appear when dangerous objects have to be handled [1] or/and when the environment is too aggressive for humans. Typical applications belong to the nuclear domain (for instance in the dismantling of a nuclear plant), deep-sea domain (work on underwater structures of oil rigs) and spatial domain (exploration of distant planets).

Teleoperation has the supplementary advantage of giving the possibility of sharing an experiment between several operators located in distinct places. This way, heavy outdoor experimentations could be easily shared between several laboratories and costs could be reduced as much. However, long distance control of a remote system requires the use of different transmission media which causes two main technical problems in teleoperation: limited bandwidth and transmission delays due to the propagation, packetisation and many other events digital links may inflict on data [2]. Moreover bandwidth and delays may vary according to events occurring all along the transmission lines. In acoustic transmission, round-trip delays greater than 10s and bit-rates smaller than 10kbits/s are common.

These technical constraints result in one hand in difficulties for the operator to securely control the remote system and, in the other hand, make classical controls unstable. Many researches have proposed solutions when delays are small or constant (for instance [3]), but when delays go beyond a few seconds and vary a lot as over long distances asynchronous links, solutions not based on teleprogramation [4] are fewer because such delays make master and slave asynchronous and the control unstable.

Work done

We worked on a low-level teleoperation technique when transmissions go through asynchronous computer networks such as local IP networks and the Internet when necessary. This architecture provides constant delays and synchronous data transmission when network delays have jitter. It also features a prediction/estimation function located on the master in order to compensate for delayed feedback. As soon as this low level layer gets into steady state, an higher level teleoperation scheme can be used (Position Error Based, 4 channel, wave transformations, ... ) such as in [3].

This work has been applied on an enhanced mobile manipulator (see [1])

A small piece of bibliography related to this topic

  1. A. LELEVÉ, P. FRAISSE, P. DAUCHEZ & F. PIERROT, "An enhanced mobile manipulator”, Proc. of the 4th World Automation Congress (WAC 2K), Maui, Hawaii, USA, 2000.
  2. R. OBOE, P. FIORINI, "Internet-based telerobotics: problems and approaches", Proc Intl. Conf. on Advanced Robotics (ICAR’97), Monterey (CA - USA), July 1997, pp. 765—770.
  3. G. NIEMEYER & J-J. E. SLOTINE, “Designing Force Reflecting Teleoperators with Large Time Delays to Appear as Virtual Tools”, Proc. of the IEEE Intl. Conf. on Robotics and Automation (ICRA’97), pp. 2212—2218.
  4. P. G. BACKES, K. S. TUO, J. S. NORRIS, G. K. THARP, J. T. SLOSTAD, R. G. BONITZ & K. S. ALI, “Internet-Based Operations for the Mars Polar Lander Mission”, Proc. of the IEEE/ASME Intl. Conf. on Advanced Intelligent Mechatronics, September 1999, pp. 317—322.