Future Is Now

Today Artificial Intelligence (AI) in Manufacturing is inspiring a revolution! It is no longer a distant vision but a miraculous reality that is being promoted by many innovation leaders at GlobalFoundries Singapore (GFS). Here is a story of how SixSense embarked on this journey of AI when GFS connected with SixSense at the SSIA Automation Supplier Day 2019 and initiated a paradigm shift in digitisation.

Collaboration with SixSense for Enabling AI-based Industrial Visual Inspection

GFS is adopting a new technology called classifAI, an AI-based Auto-Defect Classification software, that is transforming the way industrial visual inspection is performed today. It has been developed by SixSense, a Singapore-based startup promoting smarter manufacturing in the semiconductor industry. SixSense has combined their expertise in deep learning with the insights of the inspection specialists to build a product that takes factory automation to the next level. With deployments planned to span across fabs, and globally, this could change the way we review and classify defects.

Manual Image Classification at Wafer Inspection Stations

“Currently, visual defect review and classification are conducted 24×7 by highly trained human operators who closely examine images through the naked eye,” said Mr Kian Huat Ang, Manager of Yield Defect Density at GFS. As chip sizes continue to shrink, manufacturing complexity is sharply increasing, making identification of defects extremely challenging. The repetitive and cognitively demanding nature of manual classification has to change and there is an imminent need for an alternative solution.

An AI-powered Future for Wafer Inspection

 “ClassifAI is addressing such challenges by reducing dependency on human judgements, preserving valuable domain knowledge and adding consistency to processes,” said Ms. Avni Agrawal, Co-founder and CTO at SixSense.

The technology is designed to handle high volume and broad variety of complex manufacturing defects at GFS. The software learnt visual patterns of defects from images of various inspection layers, devices and technology nodes. It now classifies defects in new images from production at high speed and accuracy. With its auto-learning capability, it constantly adapts to changes in defects, making the engine more dynamic and robust.

Ms Akanksha Jagwani, Co-founder and CEO at SixSense, shared: “We are committed to making the journey of AI for manufacturers smooth and sustainable using a UI-assisted continuous performance improvement framework.”

“Our team has made many efforts to set up a dedicated technology infrastructure for AI applications and is actively working to make this initiative an enormous success,” Mr Kah Hing Ting, Integrated Manufacturing IT Manager at GFS, highlighted.

Harnessing Untapped Opportunities by Reducing Time to Market

A wide-scale deployment of AI can make remote inspections a reality and will be a great accomplishment. “The success of this solution will establish high confidence in reliability and consistency of AI-based solutions, and will drive many more smart manufacturing initiatives across the company,” Mr Zhimin Gu, VP & GM at GFS 200mm Operations, expressed.

The bar on quality standards is rising and innovators are taking strides forward in their Industry 4.0 goals. It is going to be an even more exciting journey ahead and we cannot wait to see where the AI technology will take us next!

ABOUT THE AUTHORS
Ah Choi Yap, Smart Manufacturing Solution Manager at GFS. He is overseeing the automation as well as the innovative solutions and adaptations at the 200mm Fab of GFS.

Madalin Foghis, Business Development at SixSense