PROJECT SUMMARY
A leading U.S.-based polysilicon manufacturer relied on manual, disconnected processes for scheduling and fleet operations, leading to higher energy costs and limited operational visibility.
PalTech enabled agentic, context-aware workflows, improving coordination and delivering up to 20% reduction in energy costs.
CHALLENGES
The organization faced systemic challenges common in large-scale manufacturing and supply chain operations:
- Production scheduling optimization was limited, relying on static inputs rather than real-time electricity pricing and operational constraints.
- Energy optimization opportunities were missed, leading to higher production costs despite meeting output targets.
- Fleet operations and plant data existed in silos, limiting real-time coordination across logistics and production.
- Operational decision-making was manual, requiring engineers to reconcile scheduling, cost, and asset availability.
- Lack of integrated manufacturing analytics, making it difficult to correlate operational efficiency with financial outcomes.
These gaps highlighted the need for a data-driven, AI-enabled manufacturing system that could unify scheduling and fleet operations.
SOLUTION
Making Production Scheduling Agentic
PalTech transformed the client’s scheduling process into an agentic production scheduling system designed for manufacturing cost optimization.
- Encoded production constraints (inventory, labor shifts, ramp rates, machine availability) into mathematical optimization models.
- Implemented production scheduling optimization using GAMS and IBM CPLEX for deterministic decisioning.
- Built a Python-based scheduling engine that:
- Consumes real-time electricity pricing
- Generates optimized schedules automatically
- Quantifies energy cost before execution
This enabled energy optimization in manufacturing, allowing planners to shift production to lower-cost windows with precision.