Yield analysis system for industrial site management, with years of industrial analysis experience, using flexible and comprehensive analysis templates, advanced and rich data models, to help users accurately and efficiently solve yield-related issues.
Lack of Methods
Lack of systematic and scientific analysis methods starting from data dimensions to support solving yield issues and providing improvement strategies.
Low Analysis Efficiency
Traditional analysis methods and system tools have slow data extraction, long analysis times, and difficulty handling anomalies.
Poor Flexibility
Fixed analysis meets conventional anomaly analysis requirements but cannot efficiently analyze complex multi-factor issues.
Insufficient Intelligence
Mainly descriptive analysis like visual reports, with few intelligent applications for prediction and prevention.
One-stop analysis, reducing analysis time from days to hours or minutes, greatly improving anomaly resolution efficiency.
Using analysis models and prediction models to monitor quality anomalies, prevent abnormal products from flowing downstream, and improve overall product yield.
Integrated with FDC, SPC, APC, etc., to form a comprehensive advanced process control system, achieving timely anomaly interception, analysis, and feedback management loop.
Through yield analysis, production line monitoring, etc., reduce capacity loss, save manpower, and improve overall benefits.
Semiconductor material crystallization and slicing, chip manufacturing, packaging, design companies.
Panel OLED manufacturing and assembly companies.
Crystallization and slicing, photovoltaic chip manufacturing, photovoltaic module companies.
Printed circuit board industry and other industries requiring engineering data integration and analysis.
Client: Semiconductor display enterprise. Effect: A semiconductor display enterprise relied mostly on manual judgment and handling for anomalies; product anomaly factors are numerous and complex, when analyzing unknown multi-factors, analyzing all equipment, all parameters, and all feature values is inefficient, existing analysis tools cannot quickly analyze unknown multi-factors.
GETECH uses AI+BigData to solidify engineers' analysis methods, integrates MFA, KPC, Spotfire, etc., to build a unified yield intelligent platform, solving yield issues, analyzing correlations between product yield and various parameters like time, machines, parameters, feature values. One-stop yield report query and analysis, improving business efficiency; helping engineers quickly locate anomalies, improving anomaly resolution efficiency from days to hours; enabling user self-service rapid analysis modeling. Achieves convenient extraction of full process and equipment parameters, time reduced by 10 times; generates time+machine+parameter one-stop anomaly analysis reports, time reduced by about 50 times; overall yield improved by 3%; saves cost over 6 million yuan per year.
