Handling unpredicted motion in industrial robot workcells using sensor networks

Abstract

Purpose Aims to show how sensor networks can be used to effectively allow industrial robots to handle unpredicted movements within their workcells. Design/methodology/approach Presents the concept of using sensor networks, using off‐the‐shelf technologies, to sense in real time movements within the workcell, and to feed this information, in real time, to the industrial robot controller, to allow adaptation of the robot to the movements. Results of numerous experiments in intercept and grasp of moving objects using a prototype of the system are reported. Findings Finds that the prototype system operates effectively and in a robust fashion in over 1,000 trials. Research limitations/implications Provides new insight and understanding for engineers working in the area of visually guided robots. Practical implications Results are of significant value to practitioners seeking to expand the markets of industrial robots by showing how to expand the capabilities of current systems with off‐the‐shelf technology. Originality/value Introduces the concept of using sensor networks to drive the controllers of industrial robots to intercept unpredictable moving objects in their workcells. Results are of value to practitioners working to expand the capabilities of their systems.