Teach edge machine learning using Edge Impulse

The Edge Impulse University Program empowers educators from around the world to train the next generation of edge machine learning professionals.

We are currently no longer taking applications for kits. Please fill out the interest form to be added to our mailing list to learn about future promotions and academic edge AI gatherings.

Join the mailing list

Free hardware kits

10 Arduino Kits

Learning resources

Access to lesson plans and sample projects

Online training

Access to training videos

Community support

Join a Discord channel of edge ML experts

Trusted by OUR PROGRAM PARTNERS

How it works

Teach students to build advanced edge ML applications

Edge Impulse is an end-to-end development platform for
machine learning on any device

“Edge Impulse data acquisition interface for building real‑world datasets. The dashboard shows collected sample data with names, labels, timestamps, and lengths, alongside a panel to record new data with device, label, sample length, sensor, and frequency settings. A waveform preview of raw accelerometer data is displayed at the bottom right, with train/test split and data collection stats at the top.”

“Edge Impulse ‘Create Impulse’ interface for anomaly detection. Panels show configuration of time‑series data (axes accX, accY, accZ, window size, window increase, frequency, zero‑padding), spectral analysis settings, a neural network classifier block with selected spectral features, and an anomaly detection block using K‑means. On the right, output features display Mode 1, Mode 2, and Anomaly score with a green ‘Save Impulse’ button—illustrating how to design advanced algorithms visually.”

“Edge Impulse application testing interface for training and testing ML models. On the left, a scatterplot compares false activation rate (FAR) and false rejection rate (FRR), with green and blue points representing default and alternative configurations. Below, metrics show the selected configuration’s FAR and FRR. On the right, a training output panel displays logs of the search process, ending with ‘Job completed.’ A ‘Run test’ button is shown at the top.”

“Edge Impulse deployment screen titled ‘Deploy your impulse.’ The interface shows options to export an optimized model as a C++ library, Arduino library, Cube.MX CMSIS‑PACK, WebAssembly, or NVIDIA TensorRT library. A section below offers ready‑to‑go firmware binaries for development boards—illustrating seamless deployment on any edge device.”

Embedded machine learning courseware

To help you teach machine learning on embedded systems, we provide a collection of slides, reading material, videos, and project prompts organized into modules. You are welcome to download, modify, and integrate this material into your curriculum!

See Courseware

“Edge Impulse courseware interface titled ‘courseware‑embedded‑machine‑learning.’ It shows public modules for teaching, including ‘Module 1 – Introduction to Machine Learning’ and ‘Module 2 – Getting Started with Deep Learning,’ each marked as updated 3 days ago—providing downloadable slides, readings, videos, and project prompts for embedded ML education.”

Academic Paper

Edge Impulse partnered with Harvard SEAS to coauthor an academic paper detailing how Edge Impulse provides solutions to many of the challenges faced by the edge and embedded machine learning communities. You can read the abstract and full paper on arXiv.

See paper

Try Edge Impulse free

Create a keyword spotting model in 5 minutes

Start demo

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form.

Edge Impulse Studio comes to our rescue, helping the Electronics, Computing, and Control and Automation students to abstract frameworks, such as TensorFlow, to gather data and deploy trained models into embedded devices. Edge Impulse supported our students in all steps of their edge machine learning projects.

Prof. Marcelo José Rovai

Universidade Federal de Itajubá

The future of machine learning is tiny and bright. Edge impulse makes it easy to onboard learners and helps them get started with embedded ML

Prof. Vijay Janapa Reddi

Harvard University

Edge Impulse makes it incredibly easy to get started with machine learning and AI. I especially appreciate their commitment to engaging folks in industry, entrepreneurs, hobbyists, students, educators, researchers, and the like. Making their system accessible for the masses is what really impresses me the most about them.

Orlando S. Hoilett, Ph.D.

Purdue University

Having access to Edge Impulse was huge--it really helped my students better understand the big picture use case for many of the ML techniques we were studying.

LTC David M. Feinauer, PhD, PE

Virginia Military Institute

Industry Applications

Meet the growing industry
demand for edge ML

Machine learning on resource-constrained devices is creating a new field of study while gaining impressive relevance in the industry. New technologies are being developed to solve problems in a variety of areas.

Infrastructure

With Edge Impulse, governments and utility providers can more intelligently manage electric and water grids and local infrastructure.

Asset tracking & monitoring

With Edge Impulse, intelligent asset tracking is unlocking new insights in supply chain, logistics, agriculture and other industries where sensor, audio and vision data can be used to provide more contextual awareness to assets.

Agriculture

With Edge ML, agriculture solutions developers can help farmers automate and maximize output, leading to operational automation and machine-led decision making.

Stay up to date

Please fill out the interest form to be added to our mailing list to learn about future promotions and academic edge AI gatherings.

Join the mailing list

Need more info?

If you have any questions about the Edge Impulse University Program, please do not hesitate to contact us.

Get started with
Edge Impulse

Get started

“Edge Impulse getting started graphic: three connected steps —‘Integrate with your workflow’ with an image of hands typing on a keyboard, ‘Unlock sensor data value’ with an image of a microcontroller board, and ‘Optimize AI for any hardware’ with an image of industrial machines.”

Teach edge machine learning using Edge Impulse

Free hardware kits

Learning resources

Online training