Defect Inspection System
The AlphaDIS defect inspection system is a comprehensive inspection solution integrated with semiconductor production wisdom. It uses a sophisticated artificial intelligence algorithm to improve production efficiency, reliably increase product yield and generate substantial production benefits for customers.
In the semiconductor production process, manual quality inspection of wafers, dies and chips is not only time and labor intensive, but subject to inconsistent standards and human error. This often leads to extensive production times without actual assurance of product yield
AlphaDIS uses AI to more thoroughly and accurately identify defects in wafers, dies and chips throughout the production process. The types of defects that can occur in each different semiconductor manufacturing process vary, and therefore detection criteria must be distinct to each product type. AlphaDIS’s AI learning allows it to perform comprehensive quality inspection at each stage of IC production, identifying and pinpointing the exact origins of all defects so that operators know exactly which machines and/or processes to target for corrective action.
AlphaDIS then classifies and documents each defect, and uses this data to continuously update and improve its own testing standards. This automatic learning capability allows AlphaDIS to increase detection accuracy at a staggering rate, therefore shortening inspection times, reducing labor costs, optimizing production efficiency, guaranteeing product quality and reliably increasing product yield.
Increase inspection efficiency
Currently, the number of components that can be inspected is entirely dependent on the number of employees assigned to inspecting individual items. With AlphaDIS’s assistance, inspectors need only reconfirm those components identified by the system as defective, and validate the process by carrying out random second inspections of components judged to be ideal. This creates the same effect of a full inspection while freeing employees to do more productive work increasing yield.
Consistent inspection quality
The degree to which defects are identified often vary based on inspectors’ differing experience and skill levels. AI insures that all inspections are standardized based on exact parameters, avoiding labor inspection errors that might arise through human error or imprecision.
Improve inspection speed
AI is faster than manual operation and not subject to operator fatigue. For example, it takes about five seconds for one person to inspect one chip’s appearance; AlphaDIS can inspect one chip in 0.3 seconds. This increased inspection speed improves production efficiency.
Defect detection and classification with continuous learning
AlphaDIS has the advantage of continuous and infinite learning capacity. The system categorizes defect types and identifies their root causes, showing customers exactly where they should focus corrective efforts and offering solutions to problems previously encountered. AlphaDIS is able to learn new classification rules as it encounters different defects, becoming increasingly powerful, accurate and consistent at improving product yield.
New defect learning procedure
Defect inspection procedure