Hands-on Lab Course

Embedded AI Lab

Embedded AI and Artificial Intelligence of Things (AIoT)

From raw sensor signals to on-device intelligence. Train, optimize and deploy ML models on microcontrollers using Edge Impulse, TinyML and real hardware.
50+ lab hours · ~self-paced 
Beginer to Advanced
Embedded developers, ML engineers and graduates entering AIoT roles.

Program Snapshot

6

MODULES

18

HANDS-ON LABS

4

CAPSTONE SPRINTS

50+

             HANDS-ON HOURS

IICT GmbH- Learning Path 

Curriculum — From Basics to Project Ready

Self-paced modules designed in two structured phases. Each module bundles guided theory, hands-on labs and measurable learning outcomes.

01

Phase 01 — Basics

Learn the Technology

Introduction to AIoT and Edge Intelligence

Understand AIoT architecture and the strategic value of edge intelligence.

LEARNING OUTCOMES

  • Understand AIoT system layers
  • Distinguish Edge AI vs Cloud AI trade-offs
  • Map tools to the AIoT pipeline

Machine Learning for Embedded Sensor Data

Turn raw sensor signals into model-ready features and lightweight ML models.

LEARNING OUTCOMES

  • Apply TinyML fundamentals
  • Perform signal preprocessing and feature extraction
  • Train and optimize models with Edge Impulse

02

Phase 02 — Experiments

Practice and Build

Sensor Integration & Data Acquisition

Build reliable acquisition pipelines from MCU sensors to labeled datasets.

LEARNING OUTCOMES

  • Master UART/I²C/SPI sensor interfaces
  • Apply best practices for sampling and data quality
  • Package datasets ready for ML training

Deploying AI on Microcontrollers

Move trained models from Edge Impulse onto MCUs and validate on hardware.

LEARNING OUTCOMES

  • Understand the embedded AI deployment workflow
  • Integrate models into MCU firmware
  • Measure inference latency and energy efficiency

Wireless Communication & Low-Power AIoT

Transmit inference data efficiently and minimize device energy.

LEARNING OUTCOMES

  • Compare BLE, Wi-Fi, LoRa, NB-IoT trade-offs
  • Configure MCU low-power modes and duty cycling
  • Design energy-efficient AIoT communication

HANDS-ON LAB SET UP & TOOLCHAIN

Real Systems. Industry-Grade Tools.

Every lab runs on the same hardware and software stack used in industrial deployments — accessed remotely through the IICT Virtual Lab or shipped as a physical kit for institutional partners.

STM32 + LSM6DSL Accelerometer

ESP32 with BLE/Wi-Fi

Edge Impulse Studio

Arduino IDE Toolchain

Grafana Analytics Stack

What You'll Be Able to Do

Understand AIoT system layers
Apply TinyML fundamentals
Master UART/I²C/SPI sensor interfaces
Understand the embedded AI deployment workflow

Self-paced

Guided Modules

Certificate of Completion

PHASE 03 — PROJECT READY

Capstone — SmartSense — AI-Powered Edge Device

Design and deploy a complete Human Activity Recognition system on STM32 — from sensor calibration through model training to on-device inference.

S1

Sensor Calibration & Data Sampling

Calibrate IMU, log time-synchronized CSV at 50 Hz.

S2

Data Collection & Preprocessing

Stream via Edge Impulse CLI; label and segment data.

S3

Model Training & Optimization

Train classifier ≥90 % accuracy; quantize for MCU.

S4

Edge Deployment & Validation

Deploy on STM32; benchmark accuracy, latency and memory.

Project Outcomes

  • End-to-end project management and technical documentation
  • Apply data acquisition, feature extraction and deployment
  • Validate edge-AI inference on real hardware

Ready to start the Embedded AI learning journey?

Enroll directly online, or request a bundled institutional / corporate license that covers cohorts, lab access and certification.