AI with Model-Based Design: Virtual Sensor Modeling

Overview

This on-demand webinar, sponsored by MathWorks and presented with IEEE Spectrum and Wiley, covers a workflow for AI-based virtual sensor modeling. It shows how AI-based virtual sensors can estimate signals that are difficult or costly to measure, such as battery state of charge in Battery Management Systems. The session demonstrates designing, training, validating, verifying, compressing, and deploying these models to embedded processors within a single environment.

Key Takeaways

The webinar presents an end-to-end workflow for AI-based virtual sensor models on embedded processors.
AI-based virtual sensors estimate signals that are difficult or costly to measure, such as battery state of charge.
The example focuses on state of charge in Battery Management Systems.
The workflow covers design, verification, compression, and deployment in a single environment.
It is an on-demand webinar sponsored by MathWorks and partnered with IEEE Spectrum.

What the webinar covers

The session presents a complete virtual sensor workflow.

›It explores how AI-based virtual sensors estimate signals that are difficult or costly to measure.
›The practical example is battery state of charge in Battery Management Systems.
›It demonstrates designing, training, validating, verifying, compressing, and deploying AI-based virtual sensor models within a single environment.

The session shows how AI models can be integrated into system-level design and validated against performance, resource, and deployment constraints.

What you will learn

The webinar lists specific skills covered.

›Integrate AI models into Simulink for system-level simulation and verification.
›Apply formal verification to assess neural network behavior.
›Optimize models for memory footprint and execution speed.

Attendees also learn to generate and profile library-free C code for embedded deployment and to evaluate design tradeoffs across accuracy, performance, and deployment targets.

From model to embedded deployment

The workflow targets embedded processors.

›AI models are integrated into Simulink for system-level simulation and verification.
›Models are compressed for memory footprint reduction and execution speedup.
›Library-free C code is generated from AI models and run through processor-in-the-loop tests.

Format and sponsorship

The session is an on-demand event from named partners.

›IEEE Spectrum and Wiley present the on-demand webinar.
›It is sponsored by MathWorks.
›Viewers register to create a user profile for the hub, then log in to access the content.

Frequently Asked Questions

What is the webinar about?

It presents a workflow for designing, training, validating, verifying, compressing, and deploying AI-based virtual sensor models to embedded processors within a single environment.

What is a virtual sensor in this context?

An AI-based virtual sensor estimates signals that are difficult or costly to measure, such as battery state of charge in a Battery Management System.

What practical example does the session use?

It uses estimating battery state of charge in Battery Management Systems as its practical example.

What tools and steps are demonstrated?

The workflow integrates AI models into Simulink, applies formal verification, optimizes models for memory and speed, and generates and profiles library-free C code for embedded deployment.

Who sponsors and presents the webinar?

It is an on-demand webinar presented by IEEE Spectrum and Wiley and sponsored by MathWorks.

The webinar walks through turning AI-based virtual sensors into verified, compressed models ready for embedded deployment.