A major US-based dairy manufacturer operating six processing plants across the country faced significant pressure to optimize its weekly operations. The business required a tactical planning solution that could balance three highly interdependent, complex variables: the fixed, perishable nature of raw milk supply; the operational constraints of multiple production lines and work shifts within each plant; and the need to minimize total network costs (production, inventory, and transport) while meeting nationwide customer demand over a rolling three-month horizon.
The core challenge was one of integration. The dairy industry operates under strict constraints, particularly the short shelf life of raw materials and finished products. Traditional planning tools managed production scheduling (shifts) and distribution planning (shipments) as separate processes, leading to sub-optimal decisions, high waste, and unpredictable labor needs. The client needed a single, integrated optimization model that could simultaneously plan production schedules at the shift level and determine optimal distribution flows.
The Solution: A Unified Supply Network Digital Twin
Using Sophus X and its data automation module, Dastro, the first step involved compiling a detailed supply chain digital twin. This required reconciling granular data inputs across the network:
- Production Costs and Capacity: Line-level production rates, fixed capacity constraints, changeover matrices (setup times between different SKU types), and variable labor costs associated with different shift patterns.
- Raw Material Inputs: Fixed weekly intake volumes of raw milk for each of the six plants, accounting for the inherent perishability and storage capacity limits.
- Logistics: All inter-plant and plant-to-customer transportation rates, transit times, and associated inventory holding costs.
- Demand: Weekly customer demand forecasts for all finished goods.
The Sophus X model was constructed as a Mixed-Integer Linear Program (MILP) to make binary decisions on weekly shift patterns (e.g., whether a specific production line runs two shifts versus three) while simultaneously optimizing the flow of raw materials, work-in-progress, and finished goods across the network.
The objective function was weighted heavily toward minimizing the total cost while respecting the “soft” constraint of maintaining stable, balanced line and workforce utilization.
Key Insights and Outcomes
The integrated, constraint-aware optimization provided several critical insights that the previous, siloed planning approach could not deliver:
Optimal Production Plan
The analysis revealed that due to the fixed, perishable raw material supply, certain plants were consistently running suboptimal shift patterns. The model recommended a new, dynamically adjusted weekly shift schedule that maximized the throughput of high-margin SKUs while minimizing the spoilage cost of the fixed raw material supply at each site. This led to a significant reduction in weekly raw material waste.
Trade-Off Between Transport and Production
The model identified that it was often cheaper to run a third shift at a specific plant (incurring higher labor costs) to produce a high-volume product closer to the end-customer, rather than relying on a geographically distant, lower-cost plant and incurring high transport costs for cross-country shipment. This optimization resulted in a 5-7% reduction in total weekly transport spend through better product allocation.
Forward Visibility for Labor
By optimizing shifts over a three-month horizon, the business gained predictability regarding future labor needs and utilization targets for each plant. This provided a proactive framework for the operational teams to stabilize their workforce planning and minimize the use of costly temporary labor, a critical factor in maintaining high quality and efficiency in food manufacturing.
Ultimately, the integrated shift and supply network planning model transformed a tactical scheduling challenge into a strategic decision-making process. It provided the dairy manufacturer with the speed and accuracy needed to navigate the challenges of perishable inputs and multi-site operations, achieving a lower cost-to-serve while securing service levels.
