Artificial intelligence and machine learning are revolutionizing the way that embedded software is developed. Developers traditionally hand-coded their algorithms, but machine learning is giving developers the tools to train models that run on resource-constrained devices and even write the code for them. In this course, we will explore artificial intelligence and machine learning and how they apply to microcontroller-based embedded software development.
There are no hardware requirements for this course.
Registration and Playback are located here (May require login to access)
Day 1: AI and ML for Microcontrollers
In this introductory session, we will explore the rise of AI/ML and how it impacts embedded software development. We’ll explore the types of problems that AI/ML can help developers solve, use cases, and where the technology is going. Attendees will walk away with a general overview and understand when and where to apply AI/ML in their own systems along with some of the tools to help them along the way.
Day 2: Writing Embedded Software with ChatGPT and Open.AI
Using AI/ML on the microcontroller is not the only place these technologies can help developers. AI/ML tools like ChatGPT can be used to write embedded software, too. In this session, attendees will learn how to use ChatGPT to help them develop HALs, interfaces, libraries, and even debug their code.
Day 3: Tools for Machine Learning in Microcontrollers
Selecting the right tools to train and deploy a model to an embedded system can be confusing. In this session, attendees will learn about the different tools that are available for microcontrollers to collect data, train a model, and deploy the inference to their embedded target.
Day 4: Training a Model for the STM32
In this session, attendees will learn how to use the STM32 ecosystem and tools to acquire data and train a model. We’ll dive into the details of the STM32 AI tools and how they help developers create ML inferences without being experts in machine learning.
Day 5: Deploying Machine Learning Models
In this session, we’ll talk about tips and tricks for deploying machine learning models to embedded targets. We’ll explore a few examples and provide the attendees with the knowledge they need to start developing and deploying their own models.
Course Resources
Jacob’s General Embedded Software Resources
- Sign-Up for the Embedded Bytes Newsletter here
- Book: Embedded Software Design here
- Developing Reusable Firmware – A Practical Guide to API’s, HAL’s and Drivers here
- MicroPython Projects Book here
- Jacob’s YouTube Channel – here
