For Bosch Connected Devices and Solutions GmbH, Cartesiam’s NanoEdge AI™ Studio is a natural fit as it perfectly extends the major existing IoT product line – the Cross Domain Development Kit, the XDK.
Dr.-Ing. Ando Feyh
Head of Technical Responsibility, Bosch Connected Devices and Solutions
Using Machine Learning libraries developed with Cartesiam’s NanoEdge AI Studio, we were able to anticipate behaviors that were previously difficult to detect.
Senior Principal Architect, Schneider Electric
We have used Cartesiam’s NanoEdge AI Studio to develop a very promising preventive maintenance solution on our production sites.
Executive Managing Director
The NanoEdge AI studio solution has enabled us to launch the realization of the IRMA predictive maintenance sensor, with our own teams, without having to recruit machine learning experts
Serge De Senti
We partnered with Cartesiam on a highly innovative project, involving resources in France and China, and delivered, in a record time, a highly effective predictive maintenance solution on critical motor control applications.
Advanced R&D Manager
Literally, within a couple of weeks after we bought NanoEdge AI Studio, we launched "Wales", a detector of water leakage with embedded Edge AI technology from Cartesiam
Cartesiam's NanoEdge AI Studio, through its ease of use and the artificial intelligence applied directly in each connected object, enhances our data capture expertise.
Version 2 of NanoEdge AI Studio, in particular, allows Naval Group's teams to increase the quality of production data analysis and thus take a new step forward in naval maintenance.
Head, Service Digitalization
We needed a technology for predictive maintenance that was undeniably industrially proven and with rapid implementation. After comparing several solutions on the market, we chose NanoEdge AI from Cartesiam
Bob Assistant, based on Cartesiam technology, is now installed in dozens of industrial sites across Europe and provides 24/7 predictive maintenance for thousands of equipment such as pumps, air conditioners, engines and HVAC's
Jean-Claude Le Bleis
The integration of the NanoEdge AI learning machine, directly "at the Edge", allows the deployment of systems to improve the performance of old or recent equipment that is being retrofitted
DG de Twipi Group
Cartesiam: Leader in Edge AI market, with proven industrial references.
Driving innovation from the edge.
What if appliances or machines could autonomously learn their environment, predict anomalies and understand what may go wrong ?
Thousands of pumps, motors or home appliances already do!
What about yours?
Bringing two worlds together
Making endpoint AI-ready objects is complex as it requires two separate worlds to come together: Data scientists and embedded developers.
Too many projects rely on wishing these two universes may come together and therefore rarely go beyond a Proof of Concept.
Sensors and microcontrollers are the embedded developers’ world. They thrive into their environment, yet integrating AI into their world is not an easy task.
into objects, simply, rapidly
AI is an entirely different world, where data scientists thrive into logistic regression, K means clustering or neural networks….
Training a model for creating smart devices requires datasets, data scientists, and implies cloud computing cost.
Cartesiam addressed this complexity by developing a unique technology that allows model creation and training to take place directly in the microcontroller and to be based on signals captured in the devices.
NanoEdge AI Studio is a development environment aimed at embedded developers to create a machine learning (ML) library in just a few steps and without any prior knowledge in signal processing or data science.
The output is an ML library, extremely optimized in terms of RAM (0.5Kb to 10Kb) and running on any Arm Cortex-M microcontroller.
Edge AI: Not just the inference.
The learning too.
We invented « true machine learning at the Edge ». When it leaves the factory, your device, equipped with Cartesiam’s NanoEdge AI Library, is ready to learn.
When triggered, learning will happen in the device, providing the most accurate learning, right into its unique operating environment.
For example, an accelerometer on a pump will use the vibrations of that pump, at that location to learn its behavior and then infer based on a dynamically created model. A similar device, attached to another pump in a different location will create its own model. Both models will be 100% accurate “fit for their environment”.
Edge AI: Not just the inference.
The learning too.
You may also have extensive knowledge of your machines that you would love to bring to the edge.
NanoEdge AI Studio V2 brings this capability without having to go through all the painful process of signal processing and algorithms selection.
Just load the signals into NanoEdge AI Studio and the studio will deliver the best library so you can classify signals using the classification library.