CitationDownload as .RIS
Emerald Group Publishing Limited
Copyright © 2000, MCB UP Limited
Joe Engelberger often derides service robots that look cute, cost a lot of money to create and yet cannot actually do anything. Often the reason that they cannot do anything useful is that they do not possess suitable end-effectors (grippers, hands or tools). Our theme for this issue is "Grippers and exchange mechanisms" and we take a varied look at this whole subject area.
Existing industrial grippers are usually very simple devices, often having a single axis of movement for a vice like structure that may not even have position sensing to tell you how much it is open or closed, or even whether or not it has successfully grasped a part. This level of sophistication works fine if you only need to handle one type of rigid part. However if a variety of parts need to be handled then you either need a single gripper that can accommodate all the variants, or a range of single component grippers that are interchanged either manually or automatically by a gripper exchange mechanism.
If you need to have the robot grasp a variety of different tools (e.g. screwdriver and adhesive gun) then some sort of exchange mechanism (which usually includes a simple gripper or locking mechanism) is required. Using an exchange mechanism allows standard industrial tools to be used rather than the alternative of a custom designed multi-purpose tool which would probably end up being too cumbersome to use anyway. However if you just want to pick up a variety of parts then really the last thing you want to have to do is stop to swap grippers. Doing so takes time, and the implied use of electrical and perhaps pneumatic connections, increase the chances of system failure and costs of maintenance.
It is unlikely that anyone will ever produce a truly all purpose gripper. Even our own hands suffer from limited strength, vulnerability to damage and lack of absolute precision. However developments such as the BarrettHand featured in this issue (p. 181) show very positive movements towards flexible grippers that are capable of industrial application and programmable in both mechanical configuration and open/close actuation. They are not the answer to every gripping problem, but are now sufficiently well developed that they should be considered by anyone with more than the most basic gripper requirements.