Rushing to get there

Citation

Loughlin, C. (2001), "Rushing to get there", Industrial Robot, Vol. 28 No. 6. https://doi.org/10.1108/ir.2001.04928faa.001

Rushing to get there

Rushing to get there

Our theme for this issue is "Assembly and disassembly" and we include details of some very significant developments by Ford (in the article by Anna Kochan) and what is almost certainly the world's largest robot riveter (see Robert Dale's article) as well as BMW's new Mini assembly line (see John Mortimer's article).

However, our research papers reflect the rapid developments taking place in the service sector and in particular with mobile robots. For a service robot to perform a task it needs to have various capabilities:

• •

  It needs to know what it is meant to do.

• •

  It needs to be able to get there.

• •

  It needs to be able to do the job.

"Getting there" is covered in our paper by Genci Capi et al. which is primarily concerned with energy efficiencies of biped "walking" robots. The "wheels or walk" debate has been running for some time and will no doubt continue to do so. My opinion is that we need to be able to do both, and so defending one while criticising the other is largely a waste of time. In any given application there will probably be an optimal solution. For example, in a factory or hospital environment with level floors and access corridors, the wheel is pretty hard to beat as it provides a secure support and efficient means of locomotion. However, in a home or rural environment changes in floor height and obstacles strongly favour legs.

The trick with walking is to be able to stay upright and move from A to B efficiently. And by efficiently I mean with low energy requirements. Mobile service robots need to be no bigger than roughly human proportions if they are going to share our environment, and there is not much point in being cute and athletic if your battery pack will only keep you going for a few minutes.

I know you can argue that the environment can be modified to suit a wheeled robot but my view is that, once we have developed viable energy efficient biped solutions, that wheels will become last year's mobile solution. Ask anyone restricted to a wheelchair if they would rather be able to walk if you do not agree with me.

Wheels will always have their place, but legs have a lot going for them as well.

One aspect of "Doing the job" is tackled by Ishiguro et al. (in their article: "Robovie: an interactive robot"), and in their case the "job" is communicating and interacting with humans. People communicate in very sophisticated ways that go well beyond the spoken word. Body language, eye contact and tone of presentation are key to both the true meaning of what is being said and the readiness with which the information will be received and absorbed.

I see lots of good work being done in the above areas and many others that cover such topics as navigation, obstacle avoidance and collision detection, but to my view there is one important topic that seems be being largely ignored. This is getting the robots to actually to do something.

If we are not careful we will develop brilliant robots that can climb stairs and cheerfully tell us it is raining; but so far I have seen little work being done concerned with the act of physically moving things around. All the other aspects are of course essential parts of the jig-saw, but I would like to see much more work being done on specific tasks. We need to consider what functions we would like service robots to perform and then give them the hardware and software to do them.

We had the same problem in the early days of industrial robotics. At first people thought they could do anything because they could be programmed to move wherever we wanted. The trouble was that when they got there they had neither the senses nor physical tools to do the job.

Clive Loughlin