SensInFood - Sensor integration as the key to Food Industry 4.0

The integration of high-tech sensors for product inspection offers food companies great potential for optimising their production. This project wanted to work with food companies and their technology suppliers to address the challenges of sensor integration. We explored new techniques and advanced sensor technologies to improve efficiency, quality and safety in the food industry. 


Duration:  September 2018 - November 2020


 

Context
In an export-oriented food sector, accelerated digitisation and innovation of the production apparatus is necessary to strengthen the competitive position and better respond to changing market demands. Product quality is crucial in this context. The integration of innovative sensor technologies offers great potential for in-line inspection of processes and products and optimisation of production.

It is not obvious for food companies to make an informed assessment of the technological and economic feasibility of integrating sensors into their production systems and for their specific cases. To help tackle the challenges of integration, Sirris and Flanders' FOOD launched the SensInFood collective project (VIS project).

Objective
Innovative sensor technologies offer great potential for in-line inspection of products and processes, allowing for improved production efficiency. The overall aim of the project was to overcome barriers to the integration of sensors in food production. In this way, the Flemish food industry and its technology suppliers can fully exploit the potential of in-line sensor technologies.

 

Approach
Companies identified the integration of sensors into existing production lines to measure and analyse product properties and production data as one of the key themes within Industry 4.0. It allows them to control the production process more accurately and thus maximise the yield of the process. After all, the speed, intensity or duration of the process can be adjusted to the quality of the product achieved. This ensures that the production equipment is used most efficiently, and the end product retains its maximum quality.

Resultaten

In close cooperation with food companies and technology integrators in the user group, SensInFood analysed the integration issues for sensor integration using industrial cases.

The following topics were discussed most:

  • Which (recent or future) technologies exist for inline measurement of a specific product characteristic? This included the use of hyperspectral imaging (HSI), spectroscopy (especially (N)IR and Raman), microwave technology, vision, and more future-oriented solutions for measuring properties within a product, such as TeraHz technology. Soft sensing was also discussed. This involves the quantification of product parameters by means of related more easily measurable product or process parameters, based on an underlying mathematical model and/or calibrations. 


    Figure 1: microwave based moitsure measurement

    Figure 2: Raman spectroscopy to determine chemical composition of products
  • How can production inefficiency (such as scrap, production breakdowns, etc.) be prevented by human error? This question lies outside the scope of the project but is a consideration to be made before investing in equipment and sensors to detect errors further down the line. There is a clear link to digitalisation (e.g., digital scales when weighing a recipe, linked to a recipe list and any identification tags) and operator support (e.g., step-by-step instructions or checklists when performing complex tasks). See also the Testing ground Operator Support and the project Operator 4.0.

  • How do I get knowledge from all the sensor data present in my company?

    • Manually: to gain insight into the process and the dependencies between the process parameters and product properties.
    • Automatically: to control the process automatically based on (processed) measurement signals.

      The steps needed to identify, collect, clean up (missing data, incorrectly recorded data, incorrectly calibrated sensors, ...), synchronise the data was considered it... 

      The sensor data must often be analysed in combination with other data (e.g., recipe, quality control of incoming goods, process settings...). The sensor integration project is thus interwoven with other data systems (ERP, MES...) and business departments.

      In the automation step, machines are connected so that the data can be automatically captured and visualised in a central dashboard after processing. (see also the follow-up project Connected Manufacturing)

       

  • How do I approach such an integration project, how do I estimate the costs and the benefits?
Target group 

The project was aimed at food companies (155 companies) and production technology providers that offer smart automation and digitalisation solutions to food companies: integrators, machine builders, sensor developers and technology providers (80 companies).

References

Speerhead cluster-VIS-project

    Partners

    Sirris - Flanders’ FOOD

    Project financing
     

    Timing

    01/09/2018 – 30/11/2020

    Curious for more? Keep up trough our blogposts!
    Our experts keep you up to date on all things new in the world of production automation. They stay on top of things for you and keep you informed trough bite-sized articles. 

    How can we support you?

    Want to know more about the SensInFood project ? Are you a food company or a supplier to the food industry and do you have questions about the integration of sensors in your production process or machine, or how to bring together the data from existing sensors? get in touch with us. Our Senior Engineer Machatronics, Tania Drissen, will be happy to help.