Produmar
Before AGILEHAND
At Produmar, fish sorting involves separating a mixed group of fish into different species, males and females, immature and mature fish, diseased and clean fish, etc. Some of the parameters to analyze are texture, color and size of the fish. This is often done at the same time as grading and usually needs more skilled people.
After the sorting, the fish is cut into fillets and classified into different categories. Once again, the classification is performed manually. The last production line steps concern the transformation and packaging of frozen and deep-frozen fish. Moreover, it takes a long time to train novices and even longer for them to establish experience in practice.
After AGILEHAND
The objective is to implement a semi-automated mechanism able to sort different species of fish and detect and classify the quality of the fish. To this end, AGILEHAND Solutions will be integrated for developing Intelligent manipulators with self-adaptation that are able to handle soft and deformable products. These units will enable the production line to perform complex tasks like comprehensive moving, sorting, rejecting. at the various stages of production. The use of robot manipulators will be exploited to perform the various intricate tasks. The system must accommodate for the physical properties of moved products: the weight, softness, size, and type variation. Data sensed from the conveyor belts, will be collected and processed in real-time, providing alerts to workers, contributing to a more agile and reconfigurable sorting line. Moreover, the implementation of AGILEHAND Solutions will allow Produmar to decrease the number of operator injuries and sick days by reducing their involvement in unsafe and unhealthy tasks due to the hard labor and bare work conditions in cold production environments. AI optimization algorithms will be developed to increase the level of automation of the fish production line at different management levels, from the strategic-level down to the operational-level. This will look at production scheduling and sequencing, production line parameter tuning, and production execution control for complex products with stochastic and highly variable processing and handling times. Focus will be placed on human factors, ensuring the explainability of the algorithms and considering collaborative (human-in-the-middle) approaches.Expected Improvements
- Reduce the product grading cost by automating human operations
- Decrease in the number of operator injuries by reducing their involvement in unsafe and unhealthy tasks
- Increase product variety that can be manipulated with robots
- Decrease in the number of operator injuries by reducing their involvement in unsafe and unhealthy tasks
- Reduce the number of delicate products damaged during workpiece handling
- Reduction of material shortage
- Reduction of orders backlog
- Reduce the number of delicate products damaged during sorting and packaging
- Reduction in product gripping and handling time