As part of the project, Intelligent Robotics developed a sensing concept design for a non-x-ray beef scribing system. Site trials were performed to gather sensing data. Once the data was acquired, several neural networks were trialed to find the optimal setup to accurately predict these features.
The results obtained suggest that neural networks offer a feasible sensing alternative for beef scribing to x-ray sensing. It is also possible to identify the key features required to characterise beef scribing cuts without using x-ray sending.