As process windows shrink, maintenance quality and robust equipment controls become more essential. This is further complicated by increasing pressure to reduce ramp times and realize first to market price premiums. Engineers are at the limit of managing these challenges with traditional analytics approaches.
Falkonry’s Time Series AI Cloud provides an automated approach to your engineering teams to discover, explain and classify normal and anomalous operating behaviors more quickly, from existing operational data.
Prevent production disruption from equipment downs
Improve quality of fault detection and classification
Enable faster equipment start up and ramp
Defect inspection is an expensive process with inherent delays. Using sensor data to reduce inspection rate can save fab money
Identify correlations between sensor levels and defectivity, and use these correlations to drive a more efficient defect classification and inspection plan
Reduced inspection costs and yield loss
An imbalance in gas flow can result in too many reactive species in the chamber and loss of anisotropy in the etch process
Detect etchant MFC (Mass Flow Controller) anomalies days in advance of potential failure
Convert unscheduled downtime into scheduled downtime for more efficient fab operations
Running misprocessed wafers through the entire line results in a loss of potential throughput for the factory
Use tool signals to detect misprocessed wafers subject to delamination, and scrap bad wafers early to avoid wasted processing costs
Increased yield and throughput, ability to adjust process to reduce defects
Using Falkonry AI Cloud, a leading semiconductor company was able to improve yield and drive a more efficient defect inspection sampling plan, reducing inspection costs and yield loss.
