STMicroelectronics and Cartesiam announce general availability of NanoEdge AI Studio Now optimized for STMicroelectronics development boards

read the full Global Annoucement

NanoEdge AI StudioOne Software
500 Million of Machine Learning libraries


Try it for free today !

NanoEdge AI Studio enables embedded developers to easily create machine learning libraries autonomously learning and inferring at the Edge.

What is
NanoEdge AI Studio ?


NanoEdge™ AI Studio is a software solution running on Windows 10 or Linux Ubuntu, developed by Cartesiam.

If you want to learn, infer, and predict at the Edge, directly inside the microcontroller, you are at the right place.
NanoEdge™ AI Studio will easily let you create a machine learning static library to embed in your main program running on any ARM© Cortex© M microcontroller.

When embedded in microcontrollers, it gives them the ability to locally “learn” and “understand” sensor patterns, by themselves, without the need for the user to have additional skills in Mathematics, Machine Learning, Data science, or creation of pre-trained neural network. In 4 steps, NanoEdge™ AI Studio will let you generate a static library that contains an AI model designed to gather knowledge incrementally during a learning phase to able to detect potential anomalous machine behaviors and predict them.

To help you test your machine learning library, NanoEdge™ AI Studio comes fully loaded with NanoEdge AI Emulator (Windows and Linux)

You will be able to emulate the behavior of your NanoEdge AI Library and start testing it as if it was running in your embedded application. No need to flash your code in your microcontroller, simply test your library on your PC either from the command line, or the interface provided through NEAI Studio.

The NanoEdge AI Emulator comes with each library created by NanoEdge™ AI Studio.

The emulator is compiled using the same algorithm. This means that for one library, there is one emulator.
Using this tool, you can learn and detect as you would do in your embedded application.

Learn using the nominal signals used in the Studio, or any other dataset you consider as “nominal.”
Detect using the abnormal signals used in the Studio, or any other dataset you consider as “abnormal.”

It is as simple as that.

NanoEdge AI Studio
in four easy steps

  • STEP 1
  • STEP 2
  • STEP 3
  • STEP 4

Describe your hardware environment

Choose your microcontroller type
(ARM® Cortex® M0 to M7)
Choose the maximum amount of RAM that you wish to allocate to your library
Choose the sensor type used to collect data (accelerometer, current, HAL effect, ECG, etc…)

Provide your contextual data

Even if the learning happens inside the microcontroller, regular and abnormal signals are needed to give your library search engine some context.
NanoEdge Ai Studio uses both regular and abnormal datasets to test several machine learning model’s performance against your data, along with a multitude of different (hyper) parameters and signal processing algorithms.

Test before you compile

With NanoEdge AI Emulator, you can emulate the behavior of your custom NanoEdge AI Library directly on your PC, as if the library was running on your microcontroller.

Compile and Download

Choose your compilation flags and download your final static library ready to learn and infer in your microcontroller.
To make your life easier, you also receive a Hello World Linux and Windows emulators and documentation.

NanoEdge AI Studio
Thought for developers
Command line engine
to emulate MCU
Runs on Windows 10
and Linux
Intuitive user experience
One-click UI, no
coding necessary
Easy deployment
Static library ready
to be inserted inside
your main program
Embedded developers first
No mathematics or data science skills required.
Emulators available
Test your NanoEdge AI library on your PC (Windows and Linux) before your final compile
NanoEdge AI Library
Optimized for microcontrollers
Run on every Arm Cortex
MO to M7 class MCUs
Low RAM footprint
As low as 4Kb
Low latency
16ms inference,
cortex M4 80Mhz
Static allocation
Stack preserved and
no dynamic allocation
Floating Point Unit
Compatible with FPU
Hardware MCU
Learning at the Edge
Iterative learning in 30ms
cortex M4 80Mhz
AI at the Edge
Why Cartesiam ?

Think out of the box

Unlike the traditional approach consisting in creating a static knowledge model on the server-side using labeled data and an AI Framework such as TensorFlow or others, Cartesiam’s approach is to allow the knowledge creation through dynamic learning, directly on the equipment.

The NanoEdge AI Library generated with NanoEdge AI Studio will be in charge of creating this knowledge model directly in the microcontroller and will be customized for each device.

This allows us to answer the 2two main challenges that slow down the deployment of AI projects in the industry, namely: The lack of qualified data to train your model and the lack of data scientists to create the knowledge models using AI frameworks.

Data scientists new best friend

We all know how painful it is to capture and label data in preparation of model creation.

Server-side Machine Learning projects heavily depend on a large volume of high quality labeled data. Without sufficient data, it is impossible for an “AI” to learn, and if the situation encountered does not match learned data, it simply won’t work. Gathering data is a challenge but nothing compared to classifying and labeling them, to reflect a realistic vision of the phenomenon you want to learn. It takes a lot of time, cost and requires heavily skilled data scientists. Another important feature in Cartesiam’s approach is to help Datascientist capture only relevant data using NanoEdge AI Library to sort through the data and log only the relevant one to create a static knowledge.

Business Partners

Read the latest
Testimonial Eolane
Testimonial Eolane
Cartesiam, a B2B software publisher specializing in artificial intelligence embedded in objects, and éolane, a leader in industrial services in electronics and connected solutions, announce their partnership for the launch of Bob..
Testimonial STMicroelectronics
Testimonial STMicroelectronics
What if a system could use machine learning to train models and run them on the same microcontroller? It’s, in essence, the groundbreaking accomplishment of NanoEdge AI from Cartesiam

Technology Partners

NanoEdge AI studio