Dexterous Manipulation: Measuring dexterity and flexibility of Robot Grippers
Posted on Aug 12, 2013 10:58 AM. 4 min read time
Dexterity and flexibility are the new emerging trend in robotics. With the arrival of collaborative robots and new adaptive end effectors, a major concern was put forward in the Workshop held by the National Institute of Standard and Technologies (NIST) at Automate 2013. It was about how could we characterize flexibility or dexterity among all the possible solutions. When it will be time to choose between two products, what features should we compare to make the best choice? During the discussions, some interesting points were mentioned.
Panelists were asking themselves about which metrics exist for evaluating the technologies they are facing today and what the methods that should be used are. Unfortunately, there are no answers for those questions. They concluded that the current methods are empirical and use a trial and error framework. There is no standards for robotic hand performance and grasping at the moment. They explained that the actual features evaluated for choosing the right end effector are force, size, price and cycle time. So there is no measure for dexterity but the most meaningful unit measure is the cost savings that the right end effector can provide.
Some possible solutions were also discussed. They talked about whether having a challenge like DARPA would help to develop and define dexterous end effectors. It would be an interesting solution where a manufacturing problem could be the challenge to overcome. This scenario should a have low volume with a high mix range of part handling in order to set the best criteria for dexterity.
Another possible solution for testing and characterizing flexible end effector would be to have standard methods for different grasping types. In order to accomplish this kind of test, a kit such as the SHAP test kit used for defining the ability of a human hand or upper limb prostheses, could be used. However well this kind of test works to define the dexterity of a human, it can be quite difficult for a robotic device.
For example, when the task is to grab and lift a part, a human can easily process the information and try to do it. Then, only the dexterity is evaluated. For a robotic end effector, the device would require programmed instructions that can vary a lot between different end effectors. A good way to define if one has a better dexterity would be to look at the complexity of the programs. This could measure the built-in intelligence of a device because when minimal programming is needed to accomplish the same task, this means a higher degree of intelligence.
In conclusion, there is a need to measure the dexterity of the new technologies entering the market. Collaborative robots and adaptive end effectors now have the desired characteristics for human-robot collaboration and flexible production. However, we don’t yet know how to define these measurements in order to differentiate between the different levels of dexterity. Some paths are being explored and hopefully a solution and standards will soon be set. Meanwhile, we know that flexible end effectors can be characterized by the cost savings they provide. So, take a look at our adaptive grippers and how they can make you decrease your production costs and increase your cycle time for any of your application such as: material handling, bin picking, machine loading, etc.
Related Articles: Useful Robot Dexterity for the Factory