Buddy Lee with Mesa Associates

Scott MacKenzie hosts an industrial podcast from Power Gen in Dallas, Texas, celebrating industry professionals and discussing innovations. Buddy Lee from Mesa Associates, an AI subject matter expert, shares his background in electrical engineering and reliability engineering. Buddy discusses the application of AI in asset management, particularly in detecting anomalies and predicting failures in industrial machinery. He highlights the importance of having a good data set for AI models, emphasizing the need for accurate and reliable data. Buddy also touches on the cultural challenges of implementing AI in industrial settings and the potential for AI to improve operational efficiency and reduce maintenance costs.

Action Items

• [ ] Reach out to Buddy Lee to discuss further details on Mesa's AI-powered asset management solutions.
• [ ] Explore opportunities to instrument industrial equipment and collect relevant sensor data for AI/ML model development.
• [ ] Assess current maintenance practices and failure modes to determine if an FMEA-based AI approach could provide value.

Outline

Introduction and Welcome to Industrial Talk Podcast

• Scott MacKenzie introduces the Industrial Talk Podcast, emphasizing its focus on industry professionals and their innovations.
• Scott welcomes listeners to the platform, highlighting the importance of industry professionals in solving global problems.
• The podcast is broadcasting from Power Gen in Dallas, Texas, showcasing the latest innovations in the power generation industry.
• Scott mentions the excitement of the event and the opportunities to explore new technologies on the show floor.

Introduction of Buddy Lee and His Background

• Scott introduces Buddy Lee from Mesa Associates, who has previously been a guest on the show.
• Buddy shares his background, mentioning his education at the University of Tennessee and Carson Newman College.
• Buddy discusses his degrees in church music, electrical engineering, MBA, and Masters of Engineering in reliability engineering.
• Scott notes Buddy's enthusiasm for Tennessee and his support for the University of Tennessee's football team.

Discussion on Power Generation and AI in the Industry

• Buddy talks about his positive experience at the Power Gen conference, highlighting the technical information and new equipment available.
• They discuss the advancements in small modular reactors and AI in the hydro industry.
• Scott and Buddy touch on the role of AI in driving power demand, particularly in data centers.
• Buddy explains his role as the AI subject matter expert at Mesa Associates, focusing on industrial applications.

Buddy's Thesis on AI Models and Data Sets

• Buddy shares his thesis on using AI models with limited data sets, specifically vibration analysis from a paper machine.
• He explains the process of comparing AI models to human analysts using a seven-year database of vibration data.
• Buddy highlights the accuracy of five AI models in predicting bearing failures, achieving 95% accuracy.
• Scott and Buddy discuss the differences between large language models and small data sets in AI applications.

AI Applications in Industrial Asset Management

• Buddy describes the use of Failure Modes and Effects Analysis (FMEA) templates in AI models to detect anomalies in industrial assets.
• He explains how AI can predict the efficiency of boilers and identify early signs of failures, such as bearing issues.
• Buddy emphasizes the benefits of early detection in terms of cost, labor, and supply chain management.
• They discuss the importance of having a human in the loop to validate AI predictions and ensure accurate maintenance planning.

Deployment of AI Models in Industrial Plants

• Buddy explains the rapid deployment of AI models, which can take days to a week or two, depending on the data availability.
• He addresses the trust issues and cultural challenges in implementing AI in industrial plants.
• Buddy highlights the importance of vetting data for noise and signal-to-noise ratios to ensure accuracy.
• They discuss the need for historical data and the challenges of obtaining reliable data from various sources.

Pilot Programs and Data Collection

• Buddy talks about the process of selecting a pilot asset for AI deployment, such as a hydro turbine or a specific machine.
• He explains the importance of having a FMEA and a good data set to develop effective AI models.
• Buddy emphasizes the need for continuous training and updating of AI models based on new data.
• They discuss the regional differences in failure mechanisms and the importance of tailoring AI models to specific locations.

Digitalization and Instrumentation in Industrial Plants

• Buddy discusses the ongoing efforts in digitalization and instrumentation in industrial plants, particularly in the steel industry.
• He highlights the importance of securing data and keeping it on-premises to avoid external threats.
• Buddy explains the role of IT and OT teams in managing data and ensuring its integrity.
• They discuss the potential of AI to improve operational efficiency and reduce human errors in industrial plants.

Conclusion and Contact Information

• Scott wraps up the conversation, emphasizing the importance of leveraging AI to manage assets more effectively.
• Buddy provides his contact information, including his email, phone number, and LinkedIn profile.
• Scott encourages listeners to reach out to Buddy for more information on AI applications in the industry.
• They highlight the benefits of attending events like Power Gen to stay updated on the latest innovations and technologies.