Gripping experiences

Article Type: Editorial From: Assembly Automation, Volume 29, Issue 1

Our theme for this issue includes “flexible gripping” –which basically means the design of gripping mechanisms so that they can accommodate a wide variety of parts.

Our own hands are pretty useful. In an instant, I can go from typing on a keyboard to picking up a cup of coffee and back again with hardly a thought. I can also pick up a pen and control it to write a note or sign a signature, or grasp the sleeve of my pullover to pull it back so I can see my watch. All very impressive stuff – or at least it would be if we saw a robot performing the same functions – but we do it almost without thinking.

In many ways, the action of grasping, gripping and manipulating has a lot in common with the action of looking and visual recognition. For both our vision and handling skills, it is not so much the direct specification of the elements such as finger joints and muscles and tactile senses; or the lens, iris and rods and cones, that give us our abilities, as it is thanks to our brains that perform the acts of tactile and visual recognition of 3D objects.

In other words, it is not so much the tools at our disposal that enable us to perform various tasks, as our previous experiences of the same or similar situations.

Of course, we can hit physical limits as well. Our eyes have a finite resolution and our fingers struggle, even with tweezers, to pick up the latest miniaturized resistors or integrated circuits.

Many industrial applications that involve gripping simply do not need any degree of flexibility at all. Only a few different objects may be grasped and specialist designs are more than able to accommodate them and deliver them at high speed and with great precision. However, there are also still a great number of applications where flexible gripping is an essential precursor to automation.

Manufactured and grown foods are one obvious example where natural variations prohibit a rigid approach. If you tried to use a two fingered pneumatic gripper for grasping fruit you would very soon have damaged products on your hands.

There are also numerous applications that call for the further processing of rigid manufactured parts that need more than just customized tooling. One very common example of this is where products have ended up in a jumbled heap in a bin, perhaps dropping there directly from a plastic injection moulding machine. In others, even previously aligned products may have been knocked out of position in transportation or as a result of careless handling.

Our skills as people are highly dependant on our previous experiences. Compared to other animals we take an age to grow up and gain any form of independence, but this time is well spent as we fill our brains with novel situations and their solutions.

Expert systems are not new – although their application has not so far been as widespread as might have been expected. Expert systems basically take the advice of a human expert and convert this into a set of rules that automation and control systems can follow. In addition neural networks and fuzzy reasoning can be used to give clarity to decision making for situations that do not exactly match any of the previous criteria.

Is there not therefore a case for using human expert guidance to teach solutions using available resources, rather than the alternative approach of producing ever more complex hardware?

Clive Loughlin