AI in Industry & Semiconductor Study Group

Purdue Statistics — Fall 2026

We are forming a small AI in Industry & Semiconductor Study Group beginning in Fall 2026.

The group will consist of approximately 5–6 PhD students in the Department of Statistics at Purdue University, together with faculty and an industry mentor.

The group will meet approximately twice per month and explore how statistics, machine learning, and artificial intelligence can be applied to challenging problems in modern industrial systems, with particular emphasis on semiconductor manufacturing and related areas in industrial statistics and the ASA Interest Group on Statistics for Heavy Equipment Industries (HEI).

What We Will Study

Topics will include:

• AI for industrial systems

• Semiconductor manufacturing

• Smart manufacturing

• Statistical process monitoring

• Anomaly and change-point detection

• Predictive maintenance

• Reliability and degradation

• Semiconductor yield and quality

• Virtual metrology

• Industrial time-series and sensor data

• Digital twins

• Process optimization

• Statistical and AI research related to the HEI Interest Group

The group will combine technical study, industry discussions, hands-on data analysis, and project-based research.

This is not intended to be a traditional lecture course or journal club. A central objective is to train Statistics PhD students to translate real industrial challenges into meaningful statistical and AI research problems.

Who Should Join?

Participation is limited to PhD students in the Department of Statistics at Purdue University.

We are looking for students who:

• have a strong interest in industrial applications of statistics and AI;

• are willing to learn about semiconductor manufacturing and industrial systems;

• are interested in developing new statistical methodology motivated by real industrial problems;

• are willing to work with data and participate in project-based research;

• are interested in longer-term academia-industry collaboration; and

• can commit to regular participation in the study group.

Prior knowledge of semiconductor manufacturing is not required.

Because the group will be intentionally small, active participation and continuity across semesters will be important.

Fall 2026 Format

Participants: Approximately 5–6 Purdue Statistics PhD students, faculty, and an industry mentor

Meetings: Approximately twice per month

Meeting length: About 75–90 minutes

A typical meeting may include:

• a short student or faculty presentation;

• discussion of a paper or technical topic;

• an industry-oriented case discussion;

• hands-on data analysis; and

• discussion of ongoing student projects.

Students will take turns presenting and leading discussions.

Tentative Fall 2026 Topics

1. What Is AI in Industry?
   Industrial AI, smart manufacturing, HEI, and translating industrial problems into statistical problems

2. Semiconductor Manufacturing for Statisticians
   Manufacturing processes, HBM, advanced packaging, yield, defects, and process variation

3. Industrial Data
   Sensor data, multivariate time series, images, logs, missing data, and high-dimensional data

4. Statistical Process Monitoring
   Shewhart charts, CUSUM, EWMA, multivariate monitoring, and false-alarm control

5. Anomaly Detection and Change Detection
   Statistical and machine-learning approaches to detecting abnormal industrial behavior

6. Reliability and Predictive Maintenance
   Degradation, remaining useful life, survival analysis, and maintenance decisions

7. AI in Semiconductor Manufacturing
   Fault detection, virtual metrology, yield prediction, process optimization, and defect inspection

8. Mini-Project Symposium
   Student team presentations and discussion of next research directions

Student Expectations

Each student will be expected to:

• participate regularly in meetings;

• lead at least one technical discussion or presentation;

• complete small data-analysis exercises;

• participate in a team-based semester project;

• present project results at the end of the semester; and

• maintain a small portfolio of industrial AI work.

Long-Term Vision

The Fall 2026 study group is intended to be the first stage of a multi-year effort to develop a cohort of Purdue Statistics PhD students who can contribute meaningfully to real-world industrial research.

2026–2027 — EXPLORE & ANALYZE

Build foundational knowledge in:

• industrial statistics;

• semiconductor manufacturing;

• AI and machine learning;

• reliability and quality;

• HEI-related research problems.

Students will begin working with public and simulated industrial datasets.

2027–2028 — BUILD & PREPARE

Students will work on increasingly substantial projects involving:

• process monitoring;

• predictive maintenance;

• anomaly detection;

• yield modeling;

• virtual metrology;

• industrial time series;

• statistical methodology motivated by real industrial problems.

Students will also gain experience presenting technical results to industry-oriented audiences.

Fall 2028 and Beyond — COLLABORATE

The long-term goal is to prepare advanced Statistics PhD students to participate effectively in real-world industry-sponsored research projects as new semiconductor and advanced manufacturing opportunities emerge in Greater Lafayette.

By that stage, students should be able to:

• understand an industrial problem;

• identify relevant data;

• formulate the problem statistically;

• select or develop appropriate methodology;

• evaluate practical constraints;

• communicate effectively with engineers and industry researchers; and

• contribute to rigorous and useful solutions.

EXPLORE → ANALYZE → BUILD → PREPARE → COLLABORATE

Our goal is to develop Statistics PhD students who can enter a discussion with industry engineers, understand the scientific and operational problem, formulate it statistically, and begin developing a rigorous and useful solution.

How to Apply

Interested Purdue Statistics PhD students should submit a short statement addressing the following by email to Kiseop Lee (kiseop@purdue.edu):

1. What is your current research area?

2. What statistical, computational, or machine-learning methods do you currently use?

3. Why are you interested in industrial AI, semiconductor manufacturing, or HEI-related research?

4. Are you interested in participating in the group across multiple semesters?

Prior semiconductor experience is not required.

Because enrollment will be limited to approximately 5–6 students, selection will emphasize research interest, active participation, and potential for longer-term involvement in industrial research projects.