Bridging the gap between Edge AI and Industrial Automation.

Although we could not attend in person this year, the D4CI team participated virtually in Embedded World 2024. We were particularly impressed by the advancements in: TinyML: Running machine learning models on microcontrollers (MCUs). Safety-Critical Linux: The push towards certifying Linux for safety-critical applications (SIL 2/3). These trends align perfectly with our roadmap for the coming year.

The first quarter of 2024 has been productive for the D4CI team. We have completed an extensive benchmarking study comparing the inference performance of various edge computing platforms. Key Findings NVIDIA Jetson Orin Nano: Delivers superior performance per watt for INT8 quantized models. Raspberry Pi 5: Surprisingly capable for lightweight TFLite models, offering a cost-effective solution for non-critical monitoring tasks. The full technical report will be published in our Publications section shortly. ...

We are thrilled to announce the official launch of the D4CI Research Group. Our mission is simple yet ambitious: to bridge the gap between the rapidly evolving world of Edge AI and the established standards of Industrial Automation. Over the coming months, we will be releasing open-source tools, datasets, and technical reports aimed at helping engineers deploy robust AI models on the factory floor. Stay tuned for updates!