NanoEdge artificial intelligence software for STM32 MCUs

Cartesiam has optimised its NanoEdge artificial intelligence software for ST’s Arm-based STM32 microcontroller development boards.

NanoEdge AI Studio is designed for companies that do not have expert resources in machine learning, according to Cartesiam. “Among many improvements, it also brings live data logging in NanoEdge AI Studio directly through the STM32 serial-USB port and an enhanced version of Cartesiam’s automatic data-compliance and quality-verification tool.”

ST’s Nucleo-F401RE and Nucleo-L432KC development boards have been selected as targets.

The software suite that runs on Windows 10 or Ubuntu and enables users to generate as well as validate machine learning libraries for embedded systems.

Users can select one of the above dev boards (instead of a generic Cortex architecture) and download a custom NanoEdge AI library – allowing proofs-of-concept to be built that both learn (train) and inference on the board.

Monitoring fan vibration to detect abnormal behaviour, is one example of an application amongst others, said ST.

“The software enables on-device learning of a nominal behaviour and is capable of detecting any variation of this behaviour, even in a complex and noisy environment,” according to Cartesiam.

Validating the Cartesiam libraries on these two boards is free, said ST, “and users can expect the addition of more STM32 platforms at some point in the future, thus testifying to the close and fruitful partnership between Cartesiam and ST”.

Applications are expected developing code for connected devices, household appliances and industrial machines