Iteratively Designing Neural Networks for Edge AI
| Start Time | End Time |
|---|---|
| 14 Nov 2025, 9:00 AM EST | 14 Nov 2025, 10:00 AM EST |
| 14 Nov 2025, 2:00 PM EST | 14 Nov 2025, 3:00 PM EST |
Overview
In motor electrification, integrating Embedded AI with Field Oriented Control (FOC) for Permanent Magnet Synchronous Motors (PMSMs) offers an opportunity to enhance speed performance, accuracy, and energy efficiency. However, integration poses unique challenges due to the entrenched nature of traditional control techniques and the limited resources provided by microcontroller units (MCUs).
Attendees will learn an iterative design approach to tiny neural network (TinyNN) creation. This illustrative example will be used to drive a motor while addressing limitations of a classic Proportional-Integral (PI) controller.
During this Seminar, covered topics will include:
- Generation of synthetic training data using a Simulink model
- Building and training a tiny neural network
- Network Optimization and Compression
- C code generation and deployment of the trained neural network to a target edge device for Model Based Design Simulations and Hardware In The Loop (HIL) Testing
By utilizing the combined tools and methodologies of MathWorks and STMicroelectronics, attendees will learn how to simplify the deployment of AI models to MCUs.
Through practical exercises, participants will gain insights into crafting small, efficient network architectures that can perform complex control tasks with minimal computational resources. The webinar will cover the entire workflow—from data processing and model training to code generation and deployment optimization—equipping attendees with the skills to innovate in embedded motor control using edge AI technology.
Highlights
- Iteratively designing a tiny neural network for Embedded AI deployment
- C code generation of neural networks
- Neural network compression with limited impact on prediction quality
Please allow approximately 45 minutes to attend the presentation and Q&A session. We will be recording this webinar, so if you can't make it for the live broadcast, register and we will send you a link to watch it on-demand.
About the Presenters
Reed Axman is a Senior Partner Manager where he is responsible for enabling hardware centric workflows regarding AI for companies such as STMicroelectronics, Texas Instruments, and Qualcomm. He works with our Partners and internal MathWorks teams to enable customers looking to add embedded AI capabilities to their products. He holds a master's degree in robotics and artificial intelligence from Arizona State University where he researched soft robotics for medical applications.
Joe Sanford is a Senior Application Engineer where is PhD in Electrical Engineering and Robotics from the University of Texas-Arlington has enabled him to work closely with customers in the aerospace and defense industry on data analysis, object detection and classification, and deep learning. His experience has naturally progressed to working with large data sets, creating predictive maintenance and anomaly detection machine learning models and then deploying those models to embedded, resource limited systems.
Product Focus
We will not sell or rent your personal contact information. See our privacy policy for details.
You are already signed in to your MathWorks Account. Please press the "Submit" button to complete the process.
