Applications of knowledge based reasoning to natural product processing

Natural products, such as meat, fish and wood based products, exhibit a high degree of variability. Key factors for successful automation of the processing of these products are the ability to sense workpiece variations and the flexibility to alter control strategies accordingly. The paper describes two applications which combine a sensor-equipped robot system with knowledge based control for the processing of naturally variable products. The first application involves the removal of lateral pin bones from fish fillets, and the second application involves optimised packing of the fish fillets. The success of these two applications at handling the natural variability of these natural products demonstrates the viability of this approach.