MicroEJ, a provider of software containers for Internet of Things (IoT) and embedded devices, and Au-Zone Technologies, an AI specialist, are now partners to accelerate the adoption of AI at the edge.

While we’re not unfamiliar with AI tools and models for edge devices, their popularity has been thwarted by the complexity of integrating them into existing applications. The difficulties are merging existing code with machine learning code, sharing computing resources, and “Update live neural networks without disrupting operations”said MicroEJ. Thanks to their rapprochement, the latter and Au Zone claim to solve these problems, above in particular, Au-Zone’s DeepView engine and toolbox: they guarantee developers a graphical development environment “transparent and intuitive” to design, train, and deploy neural networks optimized for the edge. MicroEJ complements Canadian Au Zone expertise by providing application containers “light and flexible” that enable the use of AI on any microcontroller or application processor while reducing performance, processing power and memory requirements.

Manufacturers can bring cloud-native capabilities to Edge AI, but also benefit from proven AI models for specific vision applications such as object or person detection. Applications related to predictive maintenance and anomaly detection are also affected. The operation also aims, according to the French MicroEJ, “the risks and costs that previously limited the adoption of machine learning at the edge”.

Founded in 2001, Au-Zone Technologies is a provider of development tools, technical design services and intellectual property for the design of intelligent embedded vision products and solutions.