The Freudenberg Group is a technology group that contributes to the development of its customers and of society as a whole through long-term thinking and innovation. With over 10 Business Groups distributed in 60 countries, 50,000 employees and over 170 years of history, the Group boasts cutting-edge technologies and excellent products for over 50 market segments including:
- Brake systems
- Vibration control components
- Technical fabrics
- Filters
- Chemicals
- Medical products
- IT services
- Home cleaning and laundry care systems (FHCS – Freudenberg Home and Cleaning Solutions)
FHCS (Freudenberg Home and Cleaning Solutions) is a Business Group of the Freudenberg Group and produces home cleaning and laundry care systems for the consumer market. FHCS has over 3000 employees, a global distribution network and 13 production sites in 4 continents, 8 of which are located in Europe.
FHCS is a leader in the Italian market and in the main European countries with the Vileda and Gimi brands. In Monselice, specifically, the Group produces steel, aluminum or plastic resin cloth dryers in dozens of different models. From 2022 in Monselice FHCS also manufactures ironing boards.
THE NEED
The company needed to recognize and get the accurate counting of the products loaded on the handling chain and then manually placed on special pallets.
The needs, and the critical issues, come from the past and refer to the production area. The company had an important shortfall in production regarding the counting of the painted semi-finished product, produced by the painting plant. To obtain that count, it was necessary to go backwards from the product stored in the warehouse, to have a clear understanding of the work done by the painting plant. This was a rough approach, as it was difficult to define the exact moment it was painted.
The result was also a difficult management of the painted semi-finished product, both internally and externally; part of the painted items is outsourced to third-party suppliers to finalize the external assembly of the clothesline.
Furthermore, in the Monselice plant, the company has about 28 types of racks, 29 types of semi-finished products, and numerous color variables: according to a proper production rate, about 200,000 semi-finished products are managed in about 24 hours. And this is indeed a huge amount of products and variables. According to this scenario, for the use of Computer Vision, the choice of the exact position to place cameras is essential, as each piece, then, can be declined in numerous colors.
The drying rack production process begins in the Production Department where the semi-finished products are made. In this phase the semi-finished products are raw, then they are loaded onto a chain which carries them to through different steps inside the painting plant to be cleaned from the previous processes, painted, dried, and finally unloaded in various areas of the factory. It was of paramount importance to understand exactly where to employ the technology needed to recognize not only their shape, but also the color of the objects.
“Francesco Mariotti, Npo Sistemi’s Artificial Intelligence Manager, won us over: he demonstrated passion and a suitable approach in the video analysis field. It wasn't just a question of applying a given technology, but of tailoring it in our sector. Here, the Npo Sistemi team demonstrated a winning approach right from the start. He had a scientific, analytical and proper method, fit to our needs”
Francesco Masutti, Global Senior IT Analyst Production, FHP, R. Freudenberg S.a.S
SOLUTION
In collaboration with the client company, the chosen solution combines Computer Vision with Artificial Intelligence which, thanks to 3D image acquisition, allows us to detect the passage of the semi-finished products, identify the color, then classify the object based on the type and count it, so as to obtain an updated picture of each step.
The benefits of this solution are various: it is possible to know exactly the quantities produced and then match them with the production orders to have a correct and real-time mapping of the semi-finished products.
In particular, the company obtained:
- greater efficiency in production
- easier identification of production stops
- reduction of errors
- cost and time savings
- improved and simplified warehouse management, both internally and in the shipping phase to external suppliers
The company's computer system printed production orders without clear timing, because of the absence of a control system. The added value is now the possibility to develop a synoptic that makes it possible to obtain a control panel with the exact counting of the pieces, and the status of the plant. Furthermore, the possibility of replicating, with the same technology, the employed techniques, will then allow the solution to be implemented both in the painting plant and in other company areas.
The company is also changing its management system: this solution will therefore make it possible to integrate even better with the company standards about quantity declarations which until now were more random and inaccurate.
“We faced a critical issue concerning the color black, which turned out to be the most requested color. Npo Sistemi’s team created an ad hoc session to solve it in one of our laboratories, and also to apply the solution to a semi-finished product with a color that is so difficult to identify. Today we have a complete analysis that allows us to continue with its application, creating the application part in the field".
Francesco Masutti, Global Senior IT Analyst Production, FHP R. Freudenberg S.a.S
