Teaching

Learning Semiconductor Equipment Maintenance in 3D

We developed EPI FIELD LAB to help students connect classroom concepts with the equipment they describe. The interactive 3D environment makes component locations, maintenance sequences and the purpose of each step easier to understand through a repeatable chamber-seal exercise.

What students learn

• Equipment structure: locate the process chamber, operator interface and seal, and understand how these components fit together.

• Maintenance workflow: work through handover verification, chamber access, inspection, seal replacement, installation checks and reassembly.

• Engineering reasoning: connect seal condition, clean part handling and verification with contamination control and equipment reliability.

Students navigate the equipment, inspect highlighted components and follow task prompts with progress feedback. The two views below show the initial handover check and the seal-removal stage.

AI Modeling in Semiconductor Manufacturing

The University of Texas at El Paso

Spring 2026 · MFG 5390 / ECE 5390 / MECH 4395 / MECH 5390

Fall 2026 · MFG 5390 / IE 4395

Listed by UTEP under Special Topics in manufacturing, electrical, mechanical, and industrial engineering.

This course connects semiconductor process knowledge with data engineering, predictive modeling, and manufacturing decisions. Students learn how a fab operates, what its data mean, and how AI can support engineering action.

Download Sample Lecture Handouts

Week 1 — View / Download PDF
Semiconductor Manufacturing Fundamentals for Smart Manufacturing & AI
Original file: Week1_Lecture Handout(20260910-231607).pdf

Week 2 — View / Download PDF
Manufacturing Data Engineering & Predictive Maintenance
Original file: Week2_Lecture_Handout(20260910-231608).pdf

What the course covers

• The fab as a system: wafer-to-chip manufacturing, industry value chains, yield, and production flow.

• Factory information and control: MES, equipment automation, FDC, SPC, APC, and maintenance records.

• Manufacturing data engineering: time-series and event alignment, join keys, missingness, quality checks, and contextual features.

• Equipment health: health indicators, predictive maintenance, remaining useful life, and pre-alarm decisions.

• AI-enabled process decisions: virtual metrology, root-cause analysis, digital twins, and process optimization.

• From models to systems: transfer and foundation models, data governance, validation, drift monitoring, and scalable deployment.

Sample Module · Week 1

Semiconductor Manufacturing Fundamentals for Smart Manufacturing & AI

Students trace the wafer-to-chip process, explain the roles of MES/FDC/SPC/APC, calculate core yield metrics, and connect virtual metrology, predictive maintenance, and root-cause analysis with appropriate data sources and performance measures.

Sample Module · Week 2

Manufacturing Data Engineering & Predictive Maintenance

Students align sensor traces with production and maintenance events, assess data quality, construct equipment-health indicators, and estimate remaining useful life. Interpretable ARIMA baselines are introduced alongside survival analysis and time-series foundation models.

When should an epitaxy chamber be cleaned?

A running example follows deposit buildup in an epitaxy chamber. Students identify comparable growth-step sensor segments, examine physically meaningful features, distinguish recipe shifts from equipment aging, and translate predicted remaining production runs into a maintenance-planning horizon.

Hands-on learning

Using synthetic semiconductor data, students build a predictive-maintenance workflow with join keys, data-quality checks, health-indicator construction, ARIMA model selection, remaining-life estimates, and pre-alarm logic. Extension activities address recipe context, maintenance events, and censored lifetime data.

Previous Teaching · National Cheng Kung University

Intelligent Manufacturing Systems / e-Manufacturing

Industry-oriented instruction in manufacturing analytics and automation, using case studies and team projects in yield improvement, anomaly detection, predictive maintenance, and process optimization.

Interactive Teaching Demonstration

Our semiconductor equipment-maintenance simulation complements classroom discussion with an interactive educational demonstration.

Open the maintenance simulation · ChatGPT sign-in may be required.

Digital Twins for Semiconductor Smart Manufacturing

A conceptual illustration of physical semiconductor equipment connected to its digital counterpart, highlighting equipment data, monitoring, and AI-enabled manufacturing decisions.