Prof. Donghee Lee Received the Excellence Award 

at the 10^th SK hynix Industry-University Research Project Award

▲ (from the left) Junsoo Kim (SK hynix), Prof. Donghee Lee (SKKU Dept. of Systems Management Engineering), Yongsoo Ha (SK hynix)

Prof. Donghee Lee (Department of Systems Management Engineering) received the Excellence Award at the 10^th SK hynix Industry-University Research Project Outstanding Invention Award. The Award for Outstanding Invention in Industry-University Research Project is a system that selects and awards outstanding patents among patents filed while performing research projects at universities that are undergoing industry-university cooperation with SK Hynix since 2013.

The winner of the prize, Prof. Donghee Lee, was assigned the task of "How to derive process paths for cluster defects management on wafer maps", developing an algorithm derivation technology to detect specific patterns on wafer maps and search for suspected facilities that caused specific patterns on wafer maps.

There are multiple semiconductor chips on one wafer, and when wafer fabrication is completed, inspection information of all semiconductor chips included in the wafer is expressed in image form through a probe test. The defective chip can be observed directly with this wafer map. Sometimes the defective chips show skewed distribution on a certain area of the map, indicating problems such as facility failure during processing. Since semiconductors are produced in a mass production system, detection of problems at the right time and treating it can minimize the loss.

Prof. Lee automatically classified the wafer maps using a Convolutional Neural Network; CNN, quantifying the similarity of defective modes and expressed these defects according to their similarities using colors, showing defects with high similarities in similar colors on the wafer map. Also, through the ‘pattern mining’ method, the algorithm for tracking suspected facilities was developed, helping engineers to recognize unidentified, new unique patterns which overall developed an effective method to detect suspected facilities.

Through this process, it is possible to effectively perform process paths, facility effects, and facility diagnosis by analyzing process history data that can be easily secured among a large amount of process data collected and analyzed in hundreds of process stages.

Prof. Lee revealed his aspirations saying, “It's an honor to be awarded and deep appreciation to SK hynix. While performing the task, I felt the difference between theory and reality. The process of solving them one by one, narrowing the gap, and applying them to the field was rewarding. In the future, we will continue to produce excellent research results through continuous implementation of industry-university tasks.”

To prevent COVID-19 infection, this reward ceremony was held in a way that SK hynix's organization directly visited the winner and delivered the plaque.