Back to R&D main

P.PSH.1477 - LEAP 4 Beef Automated Cutting Systems – Scott pre-production module 1 prototypes

Learnings from the successful lamb boning automation initiative are now being leveraged to design beef automation modules.

Project start date: 15 September 2023
Project end date: 31 March 2025
Publication date: 02 October 2024
Project status: Completed
Livestock species: Grain-fed Cattle, Grass-fed Cattle
Relevant regions: National
Download Report (8.1 MB)

Summary

The goal for this project is to develop stages of automation modules for use in beef boning rooms for Australasian processors. This will be based upon previous pre-production prototypes and will culminate in a demonstration event to industry that seeks feedback plus expressions of interests from processors for a future staged project.

Objectives

• Multiple stage development:
o Initially aiming to deploy a simple vision and cut path system to start with looking to find a cost-effective option that replicates human abilities.
o Further stages are to be based on cost and benefit of more advanced sensing/mechanics vs achievable accuracy improvements.
• Robotic approach to prove core concepts required to produce a production machine.
• In plant trials to prove concepts on actual product and quantities in a production environment to prove integration and considerations for production machine.

Key findings

• Both factory and site demonstrations were well received by processor JBS.
• CNN vision analysis provided better accuracy than traditional vision analysis.

Benefits to industry

Assuming the next stage development machine can improve on a manual cut product by 1mm for 0-rib product and 2mm for 2-rib product, this would result in a yield benefit of:
• $0.722 AUD per 0-rib striploin*
• $1.444 AUD per 2-rib striploin*
Using the above assumed benefits, the value of equipment utilising a 2-year payback would equate to: $991,744 AUD**
* This benefit is plant and commercial solution dependent.
** There may be further costs associated with equipment installation.

MLA action

This project represents a catalyst for MLA and AMPC to develop a clear joint plan that consolidates collective past learnings in the automated beef processing space and further allows Scott and industry to bring their long-standing expertise.

Future research

The results from stage 1 indicate less accuracy obtained than is typical from a manually cut product. Therefore, a stage 2 is proposed, to implement further sensing, enhance imitating the human operator, re-implement, train, and trial.
Objectives:
• Upgrading the sensing means on the Brooklyn stage 1 test system.
• Obtain the error from the ideal cut path data for both the manual and robotic system and report.
• Given the measured improvement versus the manual operation, propose a path to a commercial system.

More information

Project manager: Jack Cook
Contact email: reports@mla.com.au
Primary researcher: Scott Automation & Robotics Pty Ltd