Interpretation of mechanical impedance profiles for intelligent control of robotic meat processing

Mechanical processing of meat and fish for packaging and marketing may involve the removal of skin, fat, and bones, and the separation of meats of different texture. In an automated processing workcell that employs robots for such purposes, proper sensing and instrumentation would be quite crucial for fast and accurate control of processing operations. In particular, force and mechanical impedance at the interface of a robotic cutter and processed object (meat) would be of significant value. Instrumentation for direct sensing would be costly and may result in a system that is noisy, less robust and sluggish. An approach has been developed where mechanical impedance is sensed by means of a software filter that uses robot/cutter motion and actuator drive current as inputs. The resulting impedance profile has to be interpreted quickly and reliably for cutter control. This paper describes the technique of impedance sensing and high-level interpretation of impedance profiles, as developed by us and implemented in a laboratory robot. Also, a hierarchical control system is described that uses impedance measurements for intelligent control of a meat processing robot. Typical results presented here have been obtained from the laboratory robot.