Proving the Value of Advanced Predictive Manufacturing: Key Takeaways From SEMICON West 2021 Hybrid

By Bill Pierson, Chair of SEMICON West Smart Manufacturing Conference

The SEMI Smart Manufacturing Pavilion, featured at each SEMICON event, is fast becoming the place to learn how fabs are leveraging data, artificial intelligence (AI), and machine learning to improve quality, output, and revenue. The recent SEMICON West 2021 show in San Francisco was no different. With 14 live presentations and panel discussions, as well as a series of virtual recordings for on-demand viewing, the breadth of impactful content spanned the themes of connecting, sensing, and predicting.

Over the course of two days, presenters shared the latest ways fabs are securing a return on investment by using smart manufacturing approaches. Several of the sessions were standing room only, reflecting the criticality of Industry 4.0 to semiconductor manufacturing.
Smart Manufacturing Pavilion

During their session Creating a Data Ecosystem to Enable End-to-End Advanced Predictive Manufacturing presented, *Laura Matz, EMD Group chief technology officer, and Raj Narasimhan, corporate vice president for global quality at Micron Technology, provided a case study of their uncommon partnership in creating a shared data ecosystem to prove the advantages of advanced predictive manufacturing. The collaboration allowed them to see the potential for the rest of the supply chain. Their goal was to encourage others to prove new theories of smarter manufacturing through partnership projects.

Micron generates 13 terabytes of data per day and is investing heavily in AI to speed up the analysis of complicated data flows from the process technologies to produce its DRAM and NAND products. Meanwhile, EMD Group manages a complex set of material combinations and analyzing it all is cost-prohibitive.
Micron image

In seeking a better approach to drive material quality while improving cost effectiveness, Matz proposed combining data from all streams – supply chain to finished goods – and applying AI to more quickly diagnose concerns. The teams invested significant time, resources, and people to the projects, which both companies felt paid off.

Materials challenges often exist below the surface of the traditional testing of materials. However, the demand to reduce variations, improve yields, and ramp up new technology nodes quickly has never abated. Using AI and the principles of smart manufacturing, they now had the tools to meet the demand, but needed the partnerships in order to access all of the required data.

The companies agreed to test the concept and combined their data to detect the underlying issues. Data sharing was key to looking at the raw materials, as well as in-process and finished good parameters. They linked it to Micron’s fab data to identify the standout parameters that mattered most. After several learning cycles, the parameters were correlated.

“The combined learning from data that was always there, just not shared, was really powerful,” Matz said. “I have confidence that moving forward, we can really align our raw materials quality parameters and manufacturing process to the fab processes, getting us on the proactive side of things.”

“Absolutely,” responded Narasimhan in agreement. “It is always great to talk to suppliers when there is NOT a problem!”

Thanks to participating companies for sharing outcomes from the project at the SEMI Smart Manufacturing Pavilion! Look for similar case studies at all SEMI Smart Manufacturing events at global SEMICON Pavilions, Chapter Meetings, as well as the Global Smart Manufacturing Conference.

*Dr. Laura Matz was subsequently appointed CEO of Athinia.

About the Author

HeadshotBill Pierson is VP of Semiconductors and Manufacturing at KX, leading the growth of streaming data analytics in this vertical. Bill is also a chair for the SEMICON West Smart Manufacturing Conference and an active team member of the SEMI Americas Chapter. He has extensive experience in the semiconductor industry including previous experiences at Samsung, ASML and KLA. Bill specializes in applications, analytics, and control. He lives in Austin, Texas, and when not at work can be found on the rock-climbing cliffs or at his son’s soccer matches.

About The Author