Simplifies embedded AI deployment with an open-source AutoML plugin for resource-constrained microcontrollers—reducing development time, complexity, and hardware tuning.
Open Source AutoML Plugin Accelerates Edge AI
Analog Devices (ADI) has released AutoML for Embedded, an open-source plugin for Visual Studio Code aimed at accelerating the development of machine learning models on constrained edge hardware. Integrated into ADI’s CodeFusion Studio and co-developed with Antmicro, the plugin is designed to make ML deployment more accessible to embedded developers—especially those working with limited memory, compute, and power budgets.
Traditionally, embedding ML models into microcontrollers required intensive tuning, deep ML expertise, and extensive trial and error. The plugin tackles this complexity head-on by automating the entire model development pipeline. It uses a hybrid AutoML engine combining SMAC (Sequential Model-Based Algorithm Configuration) with Hyperband, intelligently allocating compute to the most promising model configurations.
The tool supports ADI’s MAX78002 and MAX32690 microcontrollers and uses the Kenning framework to abstract away hardware specifics. Developers can evaluate models on Renode-based simulations or in Zephyr RTOS environments—ensuring seamless prototyping and deployment.
Performance Metrics, Real-Time Validation, and Efficiency
AutoML for Embedded not only optimizes models for edge performance but also provides real-time benchmarking data—covering inference speed, memory footprint, and accuracy—via Kenning’s built-in analytics. This helps developers select the best trade-off for their target application.With automated validation steps, the plugin ensures final models meet strict memory and compute constraints. Developers can now compress and quantize models with confidence, without sacrificing critical performance.
Edge AI development often involves managing hardware limitations, tuning hyperparameters, compressing models, and navigating platform-specific toolchains—all while maintaining accuracy and efficiency. AutoML plugin simplifies this demanding workflow, offering an open-source, developer-friendly solution that accelerates time-to-deployment and democratizes embedded AI. By abstracting complexity and automating optimization, the company is empowering more engineers to bring intelligent, power-efficient AI features to even the smallest embedded platforms.
