P.PSH.1200 - LEAP4Beef DEXA and 3D Sensing evaluation for markers guiding TEYS Primal cut paths

Summary

In order for fully automated carcase processing systems to be successful, it is imperative that the sensing system is capable of accurately identifying all the necessary cut paths and specification markers required to place the primal cuts.

The aim of this project was to develop and validate automated methods for sensing location in a beef carcase to easily find the brisket and rib strip cuts (un-obscured cuts) and the spine cuts (obscured cuts) during processing. This will provide key technologies to automate all or part of the beef boning room.

The results of this project show that the sensing system in its current set up is capable of identifying the cut location markers in a carcase. This information is critical in specifying the scope and arrangement of the final automation process to automate all, or part, of the beef boning room.

Objectives

This project evaluated the sensing capability of the current Beef DEXA and 3D profile scanning cell platform to ensure that:

• all the required un-obstructed cut markers can be identified on the sensing output images
• 3D coordinates can be obtained for all the critical un-obstructed cut location markers
• obstructed cut markers in the DEXA image requires clarity to evaluate a predictive or alternate sensing method to determine their location in 3D space.

Key findings

• Several promising sensing technologies have been investigated and an estimation for accuracy has been produced. All of the key markers can be identified by one or more technologies but with a varying degree of accuracy. Trade-offs between accuracy, speed and cost will need to be considered when automation solutions are proposed.
• The design of the X-Ray system for lean meat, fat and bone ratio requires individual optimisation against the skeleton features.
• X-Ray is required to locate the cranial end of the brisket separation cut (as defined by AusMeat), as opposed to a camera and vision analysis technique.
• Automated sensing of all the locators to perform brisket, rib strip and spine cuts is possible with current technology and likely to be more accurate than current manual processes.

Benefits to industry

The results of this project create the opportunity to increase the value of cuts and adjoining portions of meat by more accurate cutting based off an automated process.

This research provides a basis to conduct further work to determine if a prediction algorithm or sensing method should be assessed, which would increase the level of confidence that the proposed "single scanning cell" processing concept is viable.

MLA action

Three other ongoing MLA projects (P.PSH.1199; P.PSH.1204; P.PSH.0893) are being established to mitigate the risk of knowledge gaps, which will help ensure that the $30 million investment for the beef boning automation project (P.PIP.0772) has the most successful outcome.

Future research

All options explored in this project could potentially offer working solutions, but the level of accuracy will vary. Further research that focuses on the cost-benefit analysis of the different processes would give an indication as to which strategy might offer the most sensible route forward.

Related resources

MLA publications

• LEAP4Beef - DEXA cut information translation using carcase marking and verify proposed cutting and processing methods are suitable for TEYS production, MLA final report, 27 February 2020
• LEAP 4 Beef Cut information translation, MLA final report, 27 February 2020
• LEAP 4 Beef, MLA program of work, 18 July 2020