Program manager of FlexCRAFT is Wageningen University & Research (WUR), in close cooperation with TU Eindhoven, TU Delft, University of Twente, and University of Amsterdam. The program also includes representatives from the business such as Marel Poultry and Celler Land.
4 million euros
The core task of NWO, the Dutch Organization for Scientific Research, is to provide funds for talented scientists and research facilities. Every year, NWO selects the best research programs out of a vast pool of proposals. During the annual innovation congress on November 8, the extensive research program FlexCRAFT was granted more than 2.7 million euros. Business partners, including Marel Poultry, will invest 1.3 million euros.
Robots can do it
The WUR research program will develop robots which can deal with the large variety of agro-food products, in combination with ever changing environmental conditions and tasks that are typical for the agro-food chain. This is exactly the kind of robotizing that Marel Poultry is looking for.
Of course, Marel’s RoboBatcher already marks a milestone in the automation of food processing. Other processes in the poultry value chain, however, demand for similar robot technology, even with more advanced generic capabilities in active perception, planning, control, gripping, and manipulation. FlexCRAFT can support this, especially since one of its three cognitive robot project cases will focus on poultry processing. In this case, Marel Poultry will be the leading industrial partner to participate in the creation and utilization of a new generation integrated robotics system solutions.
Self-learning
"We develop generic skills for robots to handle agro-food products with diverging shape, size, and firmness," says WUR program leader professor Eldert van Henten. “Such actions may be simple for a human being, but tough challenges for a robot. The robot needs to understand what kind of food products it perceives, what condition they are in, and how to approach and treat them. The sensors collect information and add that to their domain knowledge to create a so-called ‘world model’, comparable to the knowledge and experience that people build."
Human labor
Professor Van Henten continues, "Food production must be as hygienic, efficient, and sustainable as possible. In addition, fewer and fewer people are willing to do monotonous and heavy work in hot, humid greenhouses, or refrigerated areas where, for example, chicken products are processed. Robots can offer a solution: they can continue working for a long time and also function extremely well at low temperatures."