The growing potential of artificial intelligence (AI) and machine learning (ML) in embedded systems is driving new application solutions and products, but developing AI-based systems can be challenging.
NanoEdge AI Studio from STMicroelectronics is a ML technology that accelerates innovation, able to create an optimal ML library for an AI project based on a minimal amount of data. The data can also help develop an AI solution, as shown in this video where it collects the current and vibrometer information to train an ML model to detect anomalies in the system.
A PC-based pushbutton development studio for developers, NanoEdge AI Studio runs on Windows, Linux, or Ubuntu. The platform requires no advanced data science skills, as it can create four types of optimized tinyML libraries: anomaly detection, outlier detection, classification, and regression.
These libraries can be combined and chained to create a complete edge AI solution, with the anomaly or outlier detection to detect a problem on the equipment, classification to identify the source of the problem, and regression to extrapolate information and provide real insights.
Inputs can include vibration, pressure, sound, magnetic field, and time of flight, among others, and is able to combine several signals from multiple sensors in either a single library or concurrently with multiple libraries. Learning and inference are done by the microcontroller using the NanoEdge AI on-device learning library, which streamlines the edge AI process and significantly reduces development effort.
