Less downtime, more efficiency: Coca-Cola relies on MLnext for maximum availability Reducing unplanned downtime and maintenance costs for production facilities at Coca-Cola with AI-powered predictive maintenance.

Founded in Hong Kong in 1965, Swire Coca-Cola is one of the five largest bottling partners of The Coca-Cola Company worldwide and a leading food and beverage manufacturer. The company operates in China and other regions. Swire Coca-Cola distributes nearly 60 product categories. In collaboration with the team, we optimized key challenges in its production processes.

The high maintenance costs of rotary blow-fill monoblocks present a significant challenge because their precision-engineered components require costly replacement parts and extensive, labor-intensive repairs when unexpected failures occur. Moreover, any unplanned downtime of these critical systems disrupts production, delays deliveries, and increases the risk of market share loss, particularly during peak demand periods when customer relationships and brand reputation are most vulnerable. At the same time, there is a clear need for scalability. The customer wants any solution to be implemented efficiently across multiple production lines and manufacturing facilities, including Hangzhou, Nanning, and Hefei, without relying heavily on external support.

For core optimization, our first step was to focus on 14 motors in two production lines at the Hangzhou site, particularly in the sensitive blow-molding transition process. By combining historical operating data from different production phases and both large- and small-bottle production scenarios, an AI model was developed that detects emerging fault indicators early and accurately, helping to prevent unplanned failures before they occur.

By providing a user-friendly visual modeling tool, we enable the customer to independently create models for additional lines (Nanning, Hefei) using their own production experience as a basis. This provides a flexible, self-managed approach to preventive maintenance across multiple sites and equipment.

Grafana templates enable the visualization of multi-source plant data, provide comprehensive transparency into the operating state, and form the foundation for data-driven, intelligent production decisions.

The implementation of the AI-powered MLnext solution at Swire Coca-Cola has demonstrated clear and measurable added value for production processes. With its ability to now detect equipment failures early, the company is able to significantly reduce maintenance costs, avoid unplanned downtime, and ensure production continuity during critical peak seasons. The introduction of a visual modeling tool also enables internal teams to scale the solution independently across multiple plants, cutting implementation time in half and reducing reliance on external support. When combined with real-time data transparency through Grafana, the solution empowers data-driven decision-making and lays the foundation for future digital transformation efforts.

Feel free to contact us. We will consult with you to find the ideal solution together.