By Ramya Ravichandar
Today we announced a major advancement to edge computing for Industrial IoT (IIoT)—Machine learning. Lightning ML, the newest version of the FogHorn Lightning™ edge intelligence software platform, is the industry’s first IIoT platform with integrated machine learning capabilities and universal compatibility across all major IIoT edge systems. And it’s available today.
The edge computing and big data communities thought it wasn’t possible to apply machine learning models to real-time, streaming data at the edge. The challenge was “dirty” data. How can you clean and format the data for a machine learning model to process it when it’s coming in fast and furious from hundreds of sensors? FogHorn solved this problem through its patent-pending tiny-footprint, complex event processing (CEP) analytics engine.
“The addition of FogHorn Lightning ML is a monumental leap forward in delivering on the promise of actionable insights for our IIoT customers," according to FogHorn CEO David C. King. "In the initial launch of FogHorn’s Lightning platform, we successfully miniaturized the massive computing capabilities previously available only in the cloud. This allows customers to run powerful big data analytics directly on operations technology (OT) and IIoT devices right at the edge through our CEP analytics engine. With the introduction of Lightning ML, we now offer customers the game changing combination of real-time streaming analytics and advanced machine learning capabilities powered by our high-performance CEP engine.”
Machine Learning at the Edge
Lightning ML brings the power of machine learning at the edge in three groundbreaking ways:
It leverages existing models and algorithms
Industrial customers can seamlessly plug in and execute proprietary algorithms and machine learning models on live data streams produced by their physical assets and industrial control systems.
It makes machine learning OT-accessible
Non-technical personnel can use FogHorn’s tools to generate powerful machine learning insights without the need to rely on in-house or third party data scientists.
It runs in a tiny software footprint
Lightning ML enables complex machine learning models to run on highly-constrained compute devices such as PLCs, Raspberry Pi systems, tiny ruggedized IIoT gateways, as well as more powerful Industrial PCs and servers. Even with the addition of advanced machine learning capabilities, the complete Micro edition of the Lightning ML platform requires less than 256MB of memory footprint.
Support for ARM32
The newest version of Lightning is also systems-agnostic and supports 32-bit implementations of ARM® Cortex®-A processors, which are one of the most widely used processor types for IIoT deployments. Combined with its software miniaturization and machine learning capabilities, the Lightning platform’s very small footprint and support for ARM Cortex-A processors now makes edge intelligence available to an exponentially higher number of edge compute devices from many different hardware vendors.
“Fog computing requires a variety of different compute performance levels, all of which can be enabled by the flexible, low-power ARM architecture,” said Rhonda Dirvin, director of IoT and embedded, Business Segments Group, ARM. “FogHorn Systems’ Lightning platform supports and validates ARM-based solutions in OpenFog applications, and will enable new efficiencies and applications in the industrial edge computing space.”
With Lightning ML, FogHorn is again first to bring incredibly powerful data processing capabilities to memory-constrained industrial environments.
To learn more or schedule a demo of FogHorn Lightning ML, please contact email@example.com.