Optimized process control in the meat industry by inline quality measurement and automatic sorting of meat trimmings

An primary aim of MeatAutoSort is to develop an automated system that can sort individual meat trimmings according to fat content and weight into batches of specified target fat content. This is possible by combining fat estimates from a novel NIR system with the weight of each trimming obtained from a flow weight, followed by a grader for automatic sorting controlled by designed sorting algorithms. Existing standardizing systems are usually based on measurements of ground meat. The approach of monitoring trimmings instead of ground meat is demanding and challenging, but would be highly beneficial in terms of efficient logistics and product quality. Such a system can potentially produce batches of intact trimmings with correct fat levels, which would be a novel and very profitable development in the meat industry. The proposed system will allow removal of the current system, which is based on manual sorting of trimmings into batches, followed by labour intensive work to standardise the fat content. Inte grating this increased control and trimming-level information into the logistics of the deboning plant has substantial potential and this will be explored in terms of shifting the focus from livestock population to individual carcasses and trimmings, ther eby improving raw material usage and cutting pattern selection. The system will also enable more accurate and updated yield estimate models and calculations for carcass deboning. A conservative estimate indicates that a successful implementation of such s ystems in the Norwegian meat industry implies potential savings in the order of 60 mill NOK per year.