Smart AI Orchestrator for more efficient, sustainable and quality dairy processing

The project addressed the challenges of manual micro-planning in dairy processing, which is highly complex, time-consuming, and heavily dependent on individual expertise and real-time data interpretation. Existing planning tools were unable to adapt to actual production situations or optimize the consumption of resources such as energy and water.

Efficient and flexible production planning is crucial in dairy processing due to the spoilage of raw materials when not processed in time, compliance to strict quality standards, and high energy and water usage. Applying planning tools that can not accomodate unplanned situations or miss to addequatly account for business and process constraints, can lead to production inefficiencies, higher operating costs, and a greater environmental impact. By addressing these issues through digitalization and the integration of artificial intelligence, the project contributes to greater sustainability, cost efficiency, and operational resilience, which are key priorities in the modern food industry. 

Solution

The project delivered an innovative digital solution Smart AI Orchestrator, designed to optimize production scheduling in the dairy and food processing industries. This tool uses artificial intelligence approach, specifically a neural network trained on over 1,700 real production records to create efficient, data-driven production plans. It balances operational constraints like production capacity and food safety constraints with business goals, such as delivery deadlines, resulting in better production plans that consume less energy and water for the same output.

The key innovation is a hybrid planning model that combines schedules created by artificial intelligence with recorded human corrections to that plans. Planners use a simple Excel-based interface to launch the artificial intelligence tool with a single click, while retaining the ability to manually adjust plans. This approach significantly reduces planning time (by 56%), reduces the number of errors, and increases responsiveness to changing conditions.

Smart AI Orchestrator also creates measurable resource savings, including reductions in energy, gas, water, and chemical consumption worth hundreds of thousands of euros annually. It improves overall production efficiency, increases product quality, and boosts customer satisfaction.

As a result, production becomes more nimble, competitive, employee-friendly, cost-efficient, and sustainable, and the entire value chain grows more flexible.

Smart AI Orchestrator is using data/information from core digital solution. This is composed from MePIS MES and Energy Management modules and has been used in a project as founding layer to digitalize all key dairy processing resources. The core digital MePIS layer provides operational and actionable management insight and view over production processes.

Impact 

• Electricity consumption: In two operational months, measurements and optimization trend indicate yearly savings of 6,2%.
• Natural gas consumption: The estimated yearly savings to the natural gas, based on measurements from two winter operational months are 5,13%. There is a contribution from CIP savings of 3,1% and contribution from heating of estimated 8,9% per year.
• Water consumption:In two operational months, measurements and optimization trend indicate yearly savings of 3,1%.
• Consumption of Cleaning chemicals: Chemicals are part of CIP where most of the steam is used, its relative reduction is in line with steam savings of 3,1%.
• Productivity: The usage of Smart AI Orchestrator reduced weekly production planning effort from average of 1 full working day to 3,5 hours which is a decrease of 56%.

What's next? 

Following the completion of the project, we will initiate post-project commercialisation by offering the developed solution to other dairy producers, with a strategic focus on expanding its application across the broader food and process industries. Our target is to achieve a greater than 10x multiplicative impact by leveraging the adaptability of the solution across diverse industrial sectors. To support this effort, we will develop robust marketing, sales, and customer support structures that facilitate smooth adoption by external clients. 

Internally, we will roll out the solution across all relevant MC production lines to maximise operational benefits. The system will be customized and optimized for various production environments and process types, ensuring high scalability and versatility across different use cases.

To maintain and enhance the solution's competitiveness, we will continue investing in research and development. 

Key focus areas include:

• Refining AI algorithms to further improve scheduling efficiency and anomaly detection.
• Developing additional functional modules, such as predictive quality control, emissions tracking, and supply chain optimization.
• Integrating the solution with IoT platforms and advanced sensor technologies to enable deeper process visibility and real-time decision-making.

To ensure smooth and effective adoption, we will provide comprehensive training programs and ongoing user support. Additionally, user feedback will be actively collected and analysed to drive continuous, iterative improvements in system functionality and usability.

Testimonials