Sensor system promising for knife sharpness

Musculo-skeletal injuries are in the top two reported injuries in the red meat processing sector. One of the causes of this is using well used knives that are not sharp enough. 

As part of the people and culture investment portfolio, AMPC recently invested in research that developed a wearable sensor system which measures knife sharpness continuously through monitoring cutting mechanics and can indicate in real-time when a knife has become dull.

The system was trialled at a processing plant as well as in a laboratory. Linear and angular acceleration was measured together with a video recording that was used to train a machine learning model. 

AMPC Program Manager Amanda Carter said, “We observed during the trial that working using dull knives showed much more rapid and repetitive movement compared to those using a sharper knife who had more fluid and less repetitive movement.”

Through this project machine learning models were trained to identify knife accuracy during different red meat processing tasks such as boning or slicing. The accuracy produced was 90 per cent when analysing different workers doing different tasks with knives of varying sharpness. 

The research recommends that such a senor system could work together with already in place physical knife sharpness testing done through a ribbon machine. 

A range of benefits include reduced muscular injuries, cleaner more accurate cuts, and more efficient throughput. 

AMPC’s research partners were All Energy and Swinburne University of Technology. 

All Energy Process Engineer Max Barnes said, “Wearable sensors provided us with thousands of knife sharpness indications per shift without impacting comfort or workflow, enabling much faster detection of dull knives. This will help deliver faster and longer lasting outcomes to the industry."