Intelligence at the Edge.
Build, deploy, and test artificial intelligence on small, resource-constrained devices that interact with the physical world. There are 33 hands-on projects. Real hardware. Working code.
Read it now
The complete book is available through No Starch Press. Get started with the foundations and your first project today.
Who this book is for
Embedded Engineers
You know hardware. You need a clear path into AI that works on MCUs, not GPUs.
Bridge your firmware expertise to on-device intelligence with projects that run on the boards you already own.
AI/ML Practitioners
You know machine learning. Deploying to constrained devices is a different discipline.
Learn to quantise, optimise, and deploy your models where they actually run: on edge hardware with real memory and power budgets.
Makers and Hobbyists
You want to add intelligence to your projects but lack a clear starting point.
Build devices that sense, decide, and act, starting with affordable hardware and step-by-step instructions.
Students
Academic ML feels disconnected from the real world.
Ground your machine learning knowledge in deployable, testable embedded systems you can hold in your hand.
What you will build – 33 Projects
Every chapter includes complete, buildable projects with parts lists, wiring diagrams, and tested source code. Here are six that define the scope of the book.
Person Detection Using CNNs
Arduino · TensorFlow Lite Micro
Train and deploy a convolutional neural network for visual person detection on a microcontroller. Covers model architecture, training, quantisation, and on-device inference.
Orientation Using a Complementary Filter
IMU · Sensor Fusion
Fuse accelerometer, gyroscope, and magnetometer data to estimate orientation in real time. Compare complementary, Madgwick, Mahony, and Kalman filter approaches.
Robot Arm Anomaly Detection
Sensors · ML Classification
Apply machine learning to IMU sensor data to detect anomalous behaviour in a robotic arm. A practical introduction to sensor-based ML in industrial contexts.
Real-Time Audio Noise Suppression
Raspberry Pi Pico 2 · PDM Microphone
Build an RNN-based audio noise suppression system that runs in real time on a Raspberry Pi Pico 2. Includes PDM microphone interfacing, display output, and benchmarking.
Build a Hardware AI MIDI Synthesizer
RP2040/RP2350 · GAN · Custom PCB
Design and build a hardware synthesizer that uses a GAN to generate music. Covers model quantisation, PIO programming, USB MIDI enumeration, and custom PCB design across eight sub-projects.
Compressed Sensing
Adaptive Sensing Framework
Implement compressed sensing to reduce data acquisition requirements on constrained devices. Part of the book’s three-layer adaptive sensing framework.
Chapter Contents
Embedded AI GitHub Repository
Reefwing-Software/Embedded-AI
All example code, trained models, wiring diagrams, and project files from the book. Clone the repo, connect your hardware, and build. MIT licensed.
View on GitHub →PCBWay – Custom PCB Offer
Five of the projects in my book on Embedded AI are built using custom PCBs: the Noise Suppression carrier board for the Raspberry Pi Pico 2, the Pico MIDI Keyboard and VS1053 Synthesizer that form the AI music project, and the Battery Monitor and Display & Logging shields that make up the battery characterization rig. These are circuits that do not work well on a breadboard, so each was designed as a proper board with the Gerbers, schematics, and bills of materials released under the MIT license. Getting them manufactured is straightforward: all five are shared projects on PCBWay with the fabrication files already attached, so ordering takes a few clicks and boards typically arrive within one to two weeks. To make building the projects even more affordable, PCBWay is offering readers a discount.
PCBWay is offering 200 readers of Embedded AI a $10 discount on orders over $30 (one use per user). Redeem the code EmbeddedAI2026 at checkout. The code is valid until April 30, 2027 (once activated, it must be used within six months). Each board costs around $5 but shipping is roughly $25 (depending on your location), so order the boards together to minimise freight, and the combined order will clear the $30 minimum.
PCBWay – Reefwing Project Page
