5 Minutes to understand MicroEJ response to Edge AI developer’s need
Introduction
MicroEJ addresses the need for Edge AI and Machine Learning capabilities on resource-constrained microcontrollers and embedded microprocessors with two products: MicroAI and PicoAI. Each one responds to a different developer’s need.
MicroAI:
MicroAI is a lightweight Deep Learning inference engine. Developers can deploy pre-trained machine learning models into MicroEJ applications running on ultra-low-power hardware and run inference, benifiting from hardware Neural Processing Units (NPU) acceleration.
MicroAI is designed to be model-format agnostic. It can rely on different Neural Network (NN) frameworks: LiteRT (ex-TensorFlow Lite), ONNX, ExecuTorch…
PicoAI:
PicoAI brings powerful self-learning capabilities directly to embedded devices, training each device against its specific environment on real operational data with little CPU usage. It does not rely on the cloud, neural network accelerator, or pre-trained models. By learning what “normal” looks like locally, PicoAI identifies anomalies and usage patterns 1,000 times faster than traditional methods – with precision, even in dynamic and unpredictable environments.
PicoAI is powered by Boon Logic’s patented clustering technology, Nano, known for real-time, unsupervised machine learning runtimes.
At its core, it uses a highly efficient clustering algorithm that supports up to thousands of clusters per model, far outperforming conventional edge AI solutions that typically manage only a few dozen. It makes PicoAI one of the most accurate and scalable pattern recognition engines available for embedded systems.
Comparaison
To compare the two libraries, a developer uses MicroAI to execute inferences on a pre-trained model, while PicoAI is designed for on-device learning.
PicoAI enables real-time inference for pattern recognition and anomaly detection, while MicroAI is more general and can also be used for classification, temporal data analysis, and probability-based analysis.
Edge AI integration inside the MicroEJ VEE
Summary
By combining these two libraries, MicroEJ offers a comprehensive range of solutions to bring artificial intelligence to even the most constrained devices.
Going further
- Check our documentation:
- MicroAI introduction and documentation
- PicoAI documentation
- Check our videos:
- PicoAI: https://www.youtube.com/watch?v=Zx9rNgM7o0E
- Integrate and Deploy Machine Learning at the Edge with Embedded Software Containers: https://www.youtube.com/watch?v=AJsF1ot5QKY