Smart sensor for innovative industrial food process control

A great challenge in food processing is how to cope with the huge heterogeneity and quality variation in raw materials in a way that ensures high and stable end quality and a minimum of waste. One approach is to use on-line sensors to measure the essential variation so that optimal processing is achievable. This project is developing an online and non-invasive quality-measurement solution for challenging industrial food processes, where no solutions for real-time control exist today. Such a system will give the food industry a range of novel opportunities for process control. The instrument is being based on near-infrared spectroscopy and the main challenges will be to obtain sufficiently good optical signals from the interior of the samples. The light also has to penetrate outer layers such as crust, peel and skin. The system is designed to work on-line in typical process conditions, and much of the work will be done in the process lines. Novel process control concepts will be developed based on the achieved results. In this project we are concentrating on the following applications: Determination of core temperature in heat treated meat products, dry matter in whole potatoes and fat content in whole salmon. A main challenge for these applications is to measure deep into the samples and obtain good signals while the products are passing at high speed in the processing line. The participating companies are Nortura (heat-treated sausages), Findus (potatoes), Cermaq (salmon) and TOMRA Sorting Solutions who delivers optical measurement systems. SINTEF ICT is developing smart and robust optical measurement systems, while Nofima is calibrating and adapting the system to the different applications. The research team has visited each of the food companies and has good insight into the relevant processes and is aware of some of the main challenges and opportunities. Each food company has reported the required specifications for the instrument, i.e. where in the process it should be placed, what speed the products move at, the size and size variation of the products, how deeply in to the product it is desirable to measure, high/low ambient temperatures etc. All of these parameters are important in order to obtain a good design for both optics and mechanics. TOMRA contributed with optical components and SINTEF designed and built a first lab prototype (Prototype I), including the software for controlling it. This prototype was designed in particular to determine the optimum geometry for the illumination and field of view for the different products so that the most information-rich light is measured by the instrument. During end of 2016 and in 2017 systematic experiments with this prototype on all relevant products were conducted at Nofima. We obtained very good calibrations for determination of dry matter in whole potatoes, fat in whole salmon and core temperature in sausages. Measurements were performed with no contact between sample and instrument. In fact, distance between sample and instrument was about 20 cm, a distance suitable for industrial set-ups. Based on these findings, SINTEF has developed an industrial prototype based on novel optical components from TOMRA Sorting Solutions. The instrument has been extensively tested in industrial conditions in the food companies. The new prototype has the following novelties: - Measurements are done without contact with the samples, at a distance of 20 cm - Measurements are very quick, about 50 per second. - Higher spectral resolution and longer spectral range, suitable for all applications in the project. - Instrument is built in more robust waterproof construction for industrial trials. We can measure fat content in whole salmon with an accuracy of about ca ±1.2 %-points, core temperature in sausages with an accuracy of ±1.2°C, and dry matter in whole potatoes with an accuracy of about ca ±1.0 %-points. Accuracies are satisfactory for industrial use. The instrument has been successfully tested in the actual industrial lines with the companies. Business concepts has been developed on how the companies can utilize the instrument for process optimization.

The prototype instrument works well and has been successfully demonstrated in process. There are needs in the food industry for such an instrument, to monitor food quality and control production processes. There are ongoing initiatives with the aim of commercializing the instrumentation for different food process applications. When a commercial instrument is in place we can anticipate improved utilization of food raw materials, less food loss and more profitable production. The process analytical methodology is generic, and the instruments, methods and competence gained in this project can be used in several other food industries. Success in the chosen project cases (salmon, potatoes and sausages) will inspire other food companies to improve their own processes, resulting in a more sustainable and competitive food industry. The prototype instrument has already been adapted to measure critical quality features on certain types of seafood that was not part of the actual project.

A great challenge in food processing is how to cope with the huge heterogeneity and quality variation in raw materials in a way that ensures high and stable end quality and a minimum of waste. One approach is to use on-line sensors to measure the essential variation so that optimal processing is possible. This project will develop an online and non-invasive quality-measurement solution for challenging industrial food processes, where no solutions for real-time control exist today. Such a system will give the food industry a range of novel opportunities for process control. We will focus on three selected cases: -Continuous monitoring of core temperature in heat-treated foods. -On-line determination of dry matter in whole and processed potatoes -On-line determination of fat content in whole salmon Sources for value creation will be: -On-line and non-contact measurement of core temperature gives a completely new possibility to control the heat treatment process of foods. The process can be tuned so that the core temperature reaches the critical level for safety, and at the same time optimize end quality and reduce cooking loss. -Continuous determination of dry matter in potatoes enables sorting of raw material to different applications and for optimal process control, resulting in stable quality and reduced waste. -Determination of fat content in every single salmon enables sorting to different markets and retail, which will ensure added value and full utilization of the raw material variability. The instrument will be based on near-infrared spectroscopy and the main challenges will be to obtain sufficiently good optical signals from the interior of the samples. The light also has to penetrate outer layers such as crust and skin. The system must be designed to work on-line in typical process conditions, and much of the work will be done in the process lines. Novel process control concepts will be developed based on the results.