Skip to main content

AI for dairy production optimization

Optimizing production planning


Business Category
Technical Category
Constraints and satisfiability

What is the challenge that is being addressed?

Parmalat is an Italian leading company in the milk market in Italy and part of the Lactalis Group. Parmalat has nine production plants, distributed throughout Italy and the range of products includes fresh milk, microfiltered milk, UHT sterilized milk, cream, béchamel sauce, yoghurt, dessert and fruit juices. Collecchio plant has three lines for incoming milk than can unload more than fifty milk trucks each day and has a global overall stock capacity of 2,5 million litres.

The assets in the facility are the typical milk processing plants: pasteurizers, fat titrators, homogenizers, sterilizers, aseptic packaging lines, carton box machines and palletizers.

There are also other assets for the preparation of fruit juices and yoghurt, yoghurt packaging machines and plants for the preparation and mixing of bechamel and special cream (i.e., with mushrooms, salmon, etc.).

The problem is to plan the production of the finished products on the packaging lines managing to meet the market needs and the necessary resources in terms of raw material, packaging and skills of operators.

The process of scheduling packaging lines with a finite capacity has many constraints and requires a lot of specific knowledge to plan the production in an efficient way, in order to level the pace (takt time), eliminate bottlenecks and organize resources in the best possible way.

The site has also the need to reschedule the production in real-time when some variations occur: for example, a change from demand planning or a breakdown on a line or a quality issue on incoming milk. In these situations, line managers have to react very quickly and, today, it is all related to the experience and human behaviour.

What is the AI solution the project plans to implement?

Company's expectation is to create a finite capacity scheduling tool, able to optimize production sequences respecting constraints and limits of the entire production process, improving efficiency and reducing losses.

The connection with the data from the field and their analysis will offer a realistic model of the production plant, allowing to recognize in real-time the limits of the plant's production capacity and therefore to react more promptly to changes.

By using AI technologies Parmalat would like to be able to carry out predictive simulations reducing human input to a minimum. Another benefit expected, is to improve the prediction of the quality parameters of the incoming milk (especially fat and proteins) in order to conduct a more accurate and focused operation of b2b purchase and sell.


Who will help implement this solution?

This pilot is implemented within the framework of the knowlEdge Project — Towards AI powered manufacturing services, processes, and products in an edge-to-cloud-knowledge continuum for humans [in-the-loop] — a project funded by the H2020 Framework Programme of the European Commission under Grant Agreement 957331 and conducted from January 2021 until December 2023. The knowlEdge consortium consists of 12 partners from 7 EU countries, and its solution will be tested and evaluated in 3 manufacturing sectors.