ASQ Phoenix July 9, 2026 Program Meeting – Virtual
6:00pm – 8:00pm Phoenix (UTC: -7)
AGENDA: All times are Phoenix (MST):
6:00 PM – Welcome
6:05 PM – Brief Announcements
6: 15 PM – Guest Speaker Presentation
7:15 PM – Q&A and Networking
7:45 PM – Phoenix Section Business
8:00 PM – Adjourn
SPEAKER: Marilyn Padoan Wheatley
TOPIC: From Reactive to Predictive: Practical AI for Automated Visual Inspection
BIO: Marilyn Padoan Wheatley is a Senior Systems Engineer at JMP (a subsidiary of SAS), based in Arizona. She has over 15 years of experience helping engineers and scientists apply advanced analytics to improve quality and optimize processes across diverse industries worldwide.
She holds a bachelor’s degree in business and a master’s degree in economics from Eastern Michigan University, along with a Six Sigma Black Belt from Arizona State University. Marilyn also completed advanced studies in statistics at Penn State as well as machine and deep learning at the University of California, Irvine. Fluent in English, Spanish, and Portuguese, she specializes in no-code and low-code analytics workflows and is passionate about enabling quality professionals to leverage modern statistical and AI tools to solve complex problems and drive continuous improvement.
ABSTRACT: Quality control is shifting from reactive approaches to predictive strategies, and artificial intelligence is driving that transformation. Automated visual inspection—powered by computer vision and deep learning—can detect defects such as scratches, cracks, blemishes, or missing components with speed and consistency that surpass manual checks.
This presentation demonstrates that these capabilities are accessible to quality engineers without coding or data science expertise. Because defect detection models are highly context-specific, engineers’ process knowledge and quality expertise make them uniquely suited to develop and apply these solutions.
Through a practical case study using metal castings and a live demonstration, attendees will see how to build and evaluate an image classification model in a point-and-click environment. The goal: empower engineers to use AI as a tool to improve inspection accuracy, accelerate decision-making, and enhance efficiency—without replacing human judgment.
What Attendees Will Learn:
- Why deep learning matters for visual inspection
Understand how deep learning automates feature extraction from images, reducing manual measurement compared to traditional machine learning.
- The business case for moving from sampling to 100% inspection
Learn how AI can address limitations of ANSI/ASQ Z1.4 attribute sampling and reduce customer complaints tied to missed defects.
- A practical, no-code workflow for image classification
See the key steps: importing images, labeling, setting up validation, training a model, and interpreting results—all without writing code.
- How to evaluate and improve model performance
Explore metrics like validation accuracy, confusion matrix, and ROC curves, and learn how to identify misclassified images for improvement.
- Options for deployment
Discover simple ways to apply the trained model to new data and how scoring code can enable real-time defect detection on the production line.
LOCATION: via Zoom
COST: Free. Open to all.
We will be using Zoom Meeting to broadcast the February 12, 2026, ASQ Phoenix monthly Program meeting. Register in advance for this meeting:
Zoom registration link:
https://us02web.zoom.us/meeting/register/tZIod-qpqz8vEtFYBJ-Hb6hBB3gLupnqrlFg