Cymer Supplier Degradation

Manufacturers produce and sell machines to their customers. These machines have hundreds of parts and with use, the quality of these degrades. When this happens, customers experience higher machine failure rates which leads to dissatisfaction with their purchase. 

For Cymer, a company that manufactures and maintains lasers for global chip production, responding reactively to requests for replacement parts was proving expensive and inefficient. Dysfunctional machines waiting to be repaired, as well as unhappy customers, were also costing manufacturers upwards of a million dollars.

Cymer developed an Early Warning Indicator (EWI) system that could use historical data to measure unusual fluctuations and high failure rates of machine parts. 

We worked with Cymer to analyze the results using data science methodologies to detect anomalies. This data was aggregated into an intuitive visual dashboard that highlighted the risk of failure and the anomaly of each machine part.

We produced lifespan predictions from machine diagnostics and combined these with customer demand forecasts with an average 92% accuracy rate. 

The data models resulted in machine downtime being cut by more than half and led to more efficient maintenance schedules that maximized the number of replacements per service appointment.

By scheduling maintenance before the parts failed, manufacturers were able to proactively avoid out-of-use machines. Customer satisfaction increased and inventory management costs were reduced thanks to the optimization of the supply chain.

The data dashboard is also being used by procurement teams to identify low quality parts and the suppliers that provide them. What’s more, it helps them find opportunities to negotiate low prices to reflect the low-quality parts; knowledge that allows them to change suppliers if necessary.