Eric Ronco's remarks

I am involved in control of non linear systems in the department of mechanical engineering at the University of Glasgow. I have attended the robotics multidisciplinary workshop held Monday the 10/03/97 at Edinburgh University. I found the workshop really interesting and therefore thanks the organisers. However I got some comments about what I have heard during this workshop. I am rather hoping that these comments would eventually lead to an "on-line discussion".

First of all I did not find much multidisciplinarity during the presentations and discussions. As highlighted by Barbara Webb, only one side of the robotics field was represented during this workshop, the one somehow biologicaly inspired, let us call it "natural robotics". "Natural robotics" is a marginal part of robotics. Most of the applied robotics is driven by engineering theory and especially by control engineering. If I would like to be provocative I would say that "natural robotics" has rarely seen the light of the real world and thus is not the best candidate to represent robotics.

An other thing that occurs to me is that although the word "engineering" did often come up during the workshop very little seems to be known about this field. "Engineering" seems to appeal and at the same time repulse. Do the life scientists perceive engineers as the sorcerers of the 20th centuries?

I had the feeling as well that the attendee life scientists believed that engineered robots are pretty rudimentory. The demand for robots capable of achieving complex real world tasks is increasing every day. To meet this demand engineers are developing more and more sophisticated robots. In fact the difference between "natural robotics" and "engineering robotics" concerns the approach of the problem rather than its difficulty. "Natural robotics" is somehow interested in the understanding of biological systems through the developing of robots. "Engineering robotics" aims to build robots that works. These two approaches are not contradictory and therefore cooperative research could and should be carried out.

The workshop was concluded by the question: "Is the natural robotics going to supersede engineering robotics?". I do not think that this question is relevant, the question should rather be: how life scientists and engineers can cooperate to develop more efficient robots that could at the same time give biological understandings?

In summary, robotics looks to be a sort of interface between life sciences and engineering that avoids these two fields to be in direct contact. I am convinced that both these fields would gain from more interactions.

Eric Ronco

Organizers comments:

For our part, we agree with you that robotics was not presented in its generality, in fact it was difficult to present a global view on robotics (or any field) in a one-day workshop and we think that, especially for non roboticists, more time should have been spend to review what robotics really is. However, we don't think that there exists a strong division between two types of robotics, 'natural' and 'engineered', we rather see robotics as a continuous field. And the question of taking inspiration from biology is (for us) more trying to analyse interesting features found in animals, and to adapt them on a robot (legs, arms, sensors, neural control systems,etc.) in order to see if we can improve its efficiency rather than exactly replicating an animal.

We thus totally agree that one type of robotics superseding another is not a relevant question. And although the point was risen in the second panel, we don't think it was the conclusion of the workshop.

Aude Billard and Auke Ijspeert