The weakest link

Citation

Loughlin, C. (2006), "The weakest link", Industrial Robot, Vol. 33 No. 6. https://doi.org/10.1108/ir.2006.04933faa.001

The weakest link

I must confess that I usually find robotics exhibitions rather hard work and so with this in mind I was very pleased to find myself actually enjoying the Automatica exhibition in Munich earlier this year.

In my view exhibitions should be about attracting new customers with your latest developments and innovations and it is always a bit underwhelming to see exhibits performing pick and place tasks that were boring 20 years ago.

My top prize at Automatica goes to Kuka, who had a large walk-through stand that took you on a journey encompassing a diverse range of robots and applications. I am sure a few people would label it as being a bit “Hollywood” and I am sure they can do even better next time, but at least it showed variety and creativity.

Our theme for this issue is “Robot Control and Programming” and it is this aspect of robotics that is opening new applications. In this issue, we have contributions that illustrate how multiple robots and peripheral devices can all be controlled interactively. A robot has always been just one part of an automated system, as opposed to a standalone device, but it is only comparatively recently that the necessary computing power and software has become available that allows really tight real-time integration.

The other major development has been the ability to handle sensor data including tactile feedback derived from the robot axes themselves, which allows the robot to be taught by being pushed and pulled around the workspace. I am not too sure exactly how useful this will be but the ability to detect externally applied forces and react to them will be of enormous significance as robots are used more and more working in direct collaboration with people without any of the usual safety barriers. This will be bad news for the safety barrier manufacturers but I consider that we will see major growth in robot applications that require robots and people to be working together in much the same way as people do with each other.

This may be in service applications such as medical and healthcare but is also likely to be in industrial applications including some pretty heavy and potentially dangerous tasks. There was a time when robots had mean time before failure (MTBFs) measured in terms of a few hundred hours but these days tens of thousands of hours may be expected. Also when failures do occur you can reasonably expect that the failure will be controlled in a safe and un-dramatic fashion.

Given that we can control multiple highly reliable robots in a coordinated fashion, the question then arises as to whether or not the person becomes the weak link in the chain. We can write software such that we can be confident that the robot will always do as it is told and in the correct sequence, but are we being a tad optimistic if we expect people to operate to the same levels of efficiency?

Perhaps, what we need is some way of monitoring people as they work so that the computer system follows their every move, the position of each limb and appendage and knows that the person has both performed their task correctly and kept out of harms way?

Clive Loughlin