A new edge AI platform combines local processing, industrial connectivity and low-code software to deliver real-time automation without relying on constant cloud access.
Elliance has unveiled the FalconS1 Pro, an industrial edge AI platform designed to simplify the deployment of artificial intelligence in manufacturing and automation environments. Combining dedicated AI hardware with an integrated software ecosystem, the system enables real-time inference directly on-site, reducing dependence on cloud connectivity and improving response times.
At the heart of the FalconS1 Pro is an NXP i.MX8M Plus processor paired with a Hailo AI accelerator, delivering up to 26 TOPS of inference performance. By processing sensor data locally rather than transmitting it to remote servers, the platform allows industrial systems to continue operating even when internet connectivity is limited or unavailable.
The platform is aimed primarily at manufacturing and factory automation applications. Potential use cases include automated quality inspection, where cameras identify product defects as items move along production lines, and predictive maintenance, where AI analyses vibration, temperature, pressure, current and acoustic sensor data to detect early signs of equipment failure. This enables maintenance to be scheduled before faults result in costly downtime.
On the software side, FalconS1 Pro supports Yocto-based Linux and Ubuntu, while offering Docker compatibility for application deployment. Developers can also use the accompanying EdgeMind AI Studio to build AI workflows through low-code or no-code tools, alongside an AI Model-as-a-Service platform for managing optimised model deployment. Runtime applications include optical character recognition, with additional industrial software under development.
The hardware features 8GB of RAM, 128GB of onboard storage, and industrial interfaces including UART, SPI, I²C, GPIO and CAN, making it suitable for integration into existing factory equipment and embedded systems. Elliance also plans to release open-source hardware schematics, software development kits, firmware and design resources, allowing developers to customise and expand the platform for a wide range of industrial AI applications.
