Recently, STMicroelectronics released the free NanoEdge AI Studio for implementation with ST Cortex-M 32-bit microcontrollers (MCUs). NanoEdge Studio, tailored for Arm Cortex-M microcontrollers, provides a suite of machine learning (ML) and artificial intelligence (AI) libraries.
By leveraging the program, designers can efficiently integrate typically complex ML and AI libraries into C scripts. All About Wires participated in a press briefing to explore the updated NanoEdge AI Studio further.
Dr Enables More Edge Devices to Support AI
Edge AI, akin to edge computing, functions on devices situated at the edge with certain processing capabilities. Unlike centralized AI computing, edge computing often handles smaller, more specific tasks. Due to limited computing power in high-end devices, edge AI poses greater challenges. Code optimization for speed and space efficiency becomes crucial.
facial recognition using library from NanoEdge.
ST’s NanoEdge Studio comes into play in this scenario. Specifically tailored for Cortex-M microcontrollers, NanoEdge AI Studio provides a library of specialized AI modules compatible with C-code. Engineers can articulate an application, choose from a library search within the studio, and seamlessly incorporate AI into cutting-edge options. The complimentary access to the theater significantly eases deployment across diverse products.
Remi El-Ouazzane, president of ST’s MCUs and digital ICs division, expressed, “We are confident in the strength of Nano AI Studio, hence the decision to make it widely accessible to drive AI adoption.”
Under a specific licensing agreement, customers are allowed to transition their solutions to two platforms or even just an Arm Cortex device.
The Simplified Development Process with NanoEdge Studio V4.3
For over a decade, ST has been advancing the technology required to embed AI into devices. Initially, AI concepts were ported to existing STM32 MCUs, followed by integration into automotive-grade designs. To facilitate professionals in building these ports, NanoEdge Studio was introduced in 2021. The latest iteration, V4.3, streamlines deployment and enhances development automation.
According to Marco Cassis, president of ST’s digital, MEMS, and cameras group, “Designers will discover the necessary tools and resources in the ST Edge AI suite to achieve their objectives efficiently and with customized outcomes.”
“We are simplifying the process: start from a single entry point (regardless of the ST device you are working on), navigate through a wide array of devices, and utilize the suite for any ST devices supporting STM32 MCUs and MPUs, MEMS, cameras, alongside an active community for enhanced skill-building and technology implementation.”
Diagram illustrating the NanoEdge growth process.
Prior to conducting a library search in the studio, software developers can ascertain the portion of the MCU’s resources allocated for the AI modules. The company’s C code seamlessly integrates with the library, ensuring deployability. The latest version bolsters validation to instill developers’ confidence in the algorithms’ operational efficacy.
Free and Universally Applicable
Despite the soaring demand for AI development, it remains a challenging and unconventional field of study. Particularly in the resource-constrained realm of 32-bit devices, AI development poses additional hurdles. NanoEdge Studio aims to bridge this gap, expediting the deployment of edge AI applications in diverse environments. The barrier to entry for businesses deliberating the worth of high-end AI development is significantly lowered with the availability of the tool for free.
El-Ouazzane highlighted, “NanoEdge Studio is designed to facilitate the development and deployment of neural networks on any ST devices, ensuring optimal software-hardware synergy.”
All libraries are sensor-agnostic and compatible with any Arm Cortex-M STM32 device. Moreover, with a licensing agreement between the device provider and ST, the libraries can be utilized with certain non-ST Cortex-M microcontrollers. This compatibility enables developers to work with bespoke products, 140 ST development boards, and over 1,000 Cortex-M MCUs from various manufacturers.
El-Ouazzane reaffirmed that “all tools within this AI suite are entirely free for development and production,” underscoring our commitment to accelerating the adoption of AI technologies. Furthermore, as a testament to this commitment, the NanoEdge AI libraries are now accessible for any Cortex M-based MCUs under exclusive license agreements, ensuring seamless deployment on STM32 platforms.
