Automation in the food sector, a strong mix of ingredients

The food industry is one of the strongholds within the Belgian industrial landscape. Thanks to its quality, diversity and innovative approach, the sector is growing enormously and now accounts for a tenth of our total exports. This growth is reflected in employment, exports, turnover and investments. The food industry is therefore in demand for innovative technology for higher efficiency, a more sustainable approach and a higher level of automation. These companies are therefore increasingly knocking on Flanders Make's door.

Just like our traditional manufacturing companies, the food industry is struggling with a tight labour market. Finding and training people to perform certain repetitive tasks is a constant challenge. One possible answer is extensive digitisation, which can also lead to an increase in scale that is difficult to achieve with purely manual labour. By producing food more efficiently and sustainably, companies can grow, even with a tight labour market.

We at Flanders Make are therefore getting more and more requests from this sector to jointly apply our accumulated expertise to work floors in the food industry. Climate change, intensive agriculture, the reduction of pesticides and harmful emission gases are important topics to tackle and are high on the agenda. We are already working closely with companies in the agriculture industry to make cultivation and harvesting more sustainable and efficient by introducing, for example, electrification and precision farming techniques. Meanwhile, we are also introducing more new techniques of automation, improved ergonomics, training and quality control in the processing industry.

Our approach

When companies from the food sector approach us, we always start from their specific demand. Whether they are looking for solutions in quality control, work instructions, automation or something else, we always look for a suitable solution together. Together with possible technology partners in our network, we can work out a feasibility study, proof-of-concept or demonstrator. Flanders Make promotes innovation in Flanders and beyond, so as a neutral partner we look for the best solution.

We show a number of cases below, in which we demonstrate that expertise. All these companies approached us with a specific question and received a solution, fully tailored to their company.

Digital work instructions and quality control at patisserie

Puratos is a Belgian family-owned company specialising in providing ingredients and services for the patisserie, bakery and chocolate industries. In their more than 100 years of existence, they grew into a multinational company with branches in more than 80 countries and more than 10,000 employees. Investing in R&D is part of their DNA for them and ensures they remain a major player in the market, which is why they have a team of more than 1,000 researchers and technical advisors on board worldwide.

In their Puradome FoodTech Centre, together with Flanders Make, they developed a system that can both show digital work instructions on a display and perform quality control of patisserie pieces, thus partially replacing the physical presence of chef-pastry chefs. For example, a pastry chef develops a particular pastry and its preparation can be taught digitally to other pastry chefs. The system provides step-by-step instructions that visually show how a chef performed the preparation, then a vision system also goes on to check whether the steps were performed correctly. The order of steps and the correct placement of small details is important in this respect and is also evaluated and adjusted if necessary.

This demo is currently on display at their PuraDome FoodTech Centre in Groot-Bijgaarden, where they are demonstrating new and innovative techniques and technologies as part of their "Mission to Mars" programme. The unit will move to their training centre for training patisserie chefs at a later stage.

Making smoothie machines smart

Alberts is a company that produces food machines for shops, schools and businesses, among others. The machine prepares smoothies or hot soups at the touch of a button by grinding, blending and preparing a number of frozen ingredients. Users can even use a mobile app to put together customised blends according to their taste or dietary needs. This way, any organisation can offer a healthy and tasty snack or meal in a small area. The machine also cleans itself to keep maintenance minimal.

Flanders Make supported several steps of the development. First and foremost in the development of the machine itself, which obviously has to be connected to receive orders via the app, but also to be able to carry out certain checks remotely. For this, on the advice of Flanders Make, Alberts uses EtherCat, a communication protocol for real-time controlled systems. But Flanders Make was also involved in making the production processes of the machines more ergonomic. In particular, when assembling the crusher, which crushes the frozen ingredients, some operations were not ergonomic for the operator. In the ongoing research, a cobot takes over certain tasks in that assembly, for example by holding heavy parts while the operator attaches them. Alberts is also adapting certain parts in upcoming versions of the smoothie machine to facilitate production and assembly together with robots/cobots.

Ensuring food safety with spectroscopy

Waffle manufacturer Dimabel wanted to tighten its food safety monitoring even further. Instead of a regular check per production batch, they wanted to check every waffle on their production line for harmful substances, in a fully automated process, without a laboratory. With optical spectroscopy, we can do an automated check on every wafer that passes, without damaging or even touching the wafer. This increases food safety, reduces waste and lowers the manual effort for these checks.

Optimisation of logistics processes

For Euro Gijbels, a grower and distributor, we looked for ways to improve internal logistics processes. We installed Real Time Locating Systems (RTLS) sensors on objects and transport vehicles in their warehouse to track their movements throughout the building. Based on heatmaps, we were then able to calculate the optimal locations of goods to optimise movements within the building. This improved the travel time of goods to the order picking zone and made logistics processes work more efficiently.

Yield forecasting with drones

In a collaboration with pcfruit, the Fruit Growing Experimentation Centre, we developed a proof of concept for automatic drone flights in greenhouses. In this particular study, a drone flew over strawberry plants and took continuous photos. These photos were then analysed for the density of flowers on the plants. The amount of flowers gives a direct indication of how many strawberries can be expected, so yields at harvest can be predicted. In addition, we can also generate a heatmap of those flowers on the plants, to possibly adjust things like watering and extra nutrients in the places where the yield seems lower. This automated way of monitoring crops makes for more efficient cultivation and is also applicable to different types of fruit and vegetables.

Predictive maintenance with accelerometers

For Soubry, the well-known pasta producer, we teamed up to find a way to predict faults on the pasta production line. The extrusion motor was equipped with an accelerometer capable of monitoring vibrations in the system. Then, based on historical data, we analysed how and when errors might occur. Thus, the vibrations in that motor can give an early indication of when production is in danger of going wrong and preventive action can be taken. This allows the production line to continue running until a good time for maintenance, instead of shutting down at an inappropriate time.

Technologies

Business, Sustainability & Operations Management

Optimising the operations, sustainability, and value-creation of businesses.

System-of-Systems Technologies

For a seamless interaction, coordination, and efficiency across complex industrial environments.