Software startup FogHorn Systems is bringing machine learning to the edge of the Industrial Internet of Things (IIoT) with the release Tuesday of its latest big data analytics platform, Lightning ML.

The general idea behind edge computing, and the entire "fog sector," is that it's time-consuming and expensive to move data to the cloud to analyze and then send back to the edge. For many -- but not all -- use cases, weeding through data for the valuable bits and analyzing it where it originates at the edge is more efficient and lets that company respond in near real time. (See Fog Computing: The Virtualization Angle.)

It also requires an analytics platform with a much smaller footprint. FogHorn Systems says it successfully miniaturized the massive computing capabilities available in the cloud in the previous version of its Lightning platform, released in 2016, allowing its customers to run analytics on IIoT devices right at the edge via its complex event process (CEP) analytics engine. Today's update to the platform adds machine learning capabilities to the CEP engine.

FogHorn says its machine learning works with its industrial customers' existing models and algorithms; is easy to access and understand for non-technical personnel; and can run on devices as small as tiny, ruggedized IIoT gateways. The entire software platform, which can run on-premises or connected to a public or private cloud, requires less than 256MB of memory footprint. Lighting ML supports ARM32 processors, in addition to the x86-based IIoT gateways operations technology systems that the first release supported.

For more on machine learning and artificial intelligence, visit the dedicated automation content page here on Light Reading.

FogHorn, which raised $3 million in Series A funding in May, counts big names like GE, Bosch, Dell/EMC and Yokogawa amongst its current customer roster. While it doesn't serve telecom service providers, it does shares their perspective customer base in the industrial space for industry verticals like manufacturing, oil and gas, utilities and smart buildings. All are interested in helping manufacturers as they step up their investments in IIoT and seek ways to store, analyze and understand the data the IoT devices bring with them. (See Smart Cities See Fog Rolling In.)

Accenture forecasts that IIoT will add $14.2 trillion to the global economy by 2030, making now a critical time for investment, and the MPI Group studies suggest that 89% of manufacturers will increase IoT investments in the next two years. (See The Future Is Foggy – HR Report.)

— Sarah Thomas, , Director, Women in Comms