STMicroelectronics Updates NanoEdge™ AI Studio

STMicroelectronics Updates NanoEdge™ AI Studio to Simplify Machine Learning Software Development for IoT Products and Industrial Equipment

• New algorithm can better predict equipment anomalies and future behavior characteristics

• New features and STMicroelectronics development boards make industrial sensor data collection and management more practical

• Improved user interface to make machine learning software easier to implement, even for embedded developers without data science skills

China, December 17, 2021 - STMicroelectronics (ST), a global semiconductor leader serving multiple electronic applications, today released the third version of NanoEdge™ AI Studio, a machine learning development tool. This is the first major upgrade to this machine learning application software development tool since the acquisition of Cartesiam at the beginning of the year.

The launch of the new version of NanoEdge AI Studio comes at a time when AI capabilities are shifting from the cloud to the edge, a trend that allows equipment manufacturers to fundamentally improve industrial production processes, optimize manufacturing equipment maintenance costs, bring innovative features to devices that can sense and process data and operate locally, and improve communication latency and information security. Target applications for the upgraded software include IoT devices, home appliances, and industrial automation.

NanoEdge AI Studio simplifies the development of machine learning, anomaly learning, detection and classification applications on any STM32 microcontroller. The new version also includes prediction functions, such as regression algorithm library and anomaly detection algorithm library. With this development tool, users can quickly, easily and economically integrate cutting-edge machine learning functions into their devices without the need for data science expertise.

STMicroelectronics has added native software support for all STM32 development boards, eliminating the need for developers to write code for industrial sensors with new high-speed data acquisition and management capabilities. NanoEdge AI Studio software improves data security because edge computing uses local data storage and processing technology instead of transmitting and processing data in the cloud.

Customer feedback

“We had the privilege of using NanoEdge AI Studio to develop a solution for a client in the aerospace industry,” said Steve Peguet, Scientific Director of Innovation at Alten Group, an international technology consulting and engineering firm. “Drill wear or the slightest anomaly can have serious consequences in the manufacturing of expensive parts. Alten used NanoEdge AI Studio to integrate machine learning algorithms into electric drills. The solution worked so well when tested on the production line that Alten created a standard operating procedure around the technology to support predictive maintenance for its customers and commercialized these initial developments to deploy a groundbreaking prescriptive maintenance solution for drills in their factories.”

David Dorval, CEO and founder of Stimio , a company that develops Industrial Internet of Things (IIoT) solutions for rail and other industries, said: "Our key rail customers asked us for low-power, autonomous, wireless predictive maintenance solutions to increase line uptime, optimize operating costs and avoid costly downtime incidents. The role of low-power edge AI is core to our product strategy, and after comparing multiple edge AI software solutions, we chose ST's NanoEdge AI Studio to enrich our Oxygen Edge portfolio with its powerful low-power edge AI algorithms."

“NanoEdge AI helps us reduce machine learning development time and more quickly develop safe, next-generation electronic devices that improve the health and happiness of our families and loved ones,” said Deepak Arora, president and CEO of Wearable Technologies. “By running edge AI on personal devices, we can make informed decisions quickly with higher accuracy and fewer false positives.”

NanoEdge™ AI Studio V3 Key Features and Functions

Completely redesigned user interface to make it easier for non-expert users to exploit state-of-the-art machine learning libraries.

The STWIN development board supports new high-speed data acquisition and management functions, and can easily manage all industrial sensors without writing any code

Improved support for anomaly detection is particularly useful for predictive maintenance, enabling accurate prediction of wear and tear or better handling of equipment aging.

Learn normal features with small datasets or train models with new algorithms to recognize previously unseen abnormal patterns directly on STM32 MCUs.

Added regression algorithms to extrapolate data and predict future data patterns for energy management or to predict the remaining life of equipment.

Natively supports all STM32 development boards, no configuration required.