The edge AI market — that is, the market for devices that run AI applications locally, without cloud processing — is anticipated to grow substantially in the coming months. According to Deloitte, more than 750 million edge AI chips will be sold by year-end (up from a predicted 300 million in 2017), representing $2.6 billion in revenue. And by 2024, the number of chips sold could reach 1.5 billion in total.
Mountain View, California-based FogHorn hopes to lead the charge. The company, which was cofounded by Abhi Sharma, Kamesh Raghavendra, Mohan Reddy, Sastry Malladi, and T.M. Ravi in 2014 as part of the Hive incubator, offers a range of edge intelligence software for industrial and commercial applications. In anticipation of expansion and reflecting 10 times year-over-year growth in annual license bookings, FogHorn today announced a $25 million series C funding round that brings its total raised to $72.5 million.
“2019 proved to be a banner year for FogHorn for license bookings, new customer acquisition, and innovation,” said CEO David C. King. “This latest round of funding was unsolicited, but we could not turn down the favorable valuation and a new investor in LS Corp. that will support our growth, especially in East Asia. The series C round will extend FogHorn’s ability to scale and invest in continued product innovation as we deepen our portfolio of industry use cases.”
FogHorn — whose multi-million dollar, multi-year contract customers include Stanley Black Decker, Honeywell, and other top manufacturing, power and water, oil and gas, renewable energy, mining, transportation, health care, retail, smart building and city, and connected vehicle brands — offers products that enable devices to get as close to the source of streaming sensor data as possible.
To that end, the company’s VEL Complex Edge Processor (CEP) performs real-time analysis of data streams and optimizes for constrained compute environments that have limited or no connectivity, while at the same time simplifying interoperability with existing systems by handling complex pattern recognition on high frequency and asynchronous data. Furthermore, the VEL CEP can detect events in real time and reduce required memory size by more than 80% (less than 256MB), enabling closed-loop actions that result in cost savings and more efficient machine learning pre- and post-processing, cleansing, filtering, normalization, and contextualization.
The complementary browser-based FogHorn Manager lets companies deploy configurations or custom applications across thousands of edge devices and simultaneously funnel analytic expressions and data publications to a cloud storage location (e.g., Google Cloud, Amazon Web Services, or Microsoft’s Azure). The REX tool simulates sensor traffic to debug production issues in industrial IoT environments, while VIZ lets administrators visualize real-time streams to validate sensors, troubleshoot problems with input sources, and view the output of machine learning algorithms. There’s also VEL Studio, which helps with the authoring and debugging of analytic expressions and provides templates commonly used for expression creation, and a visual debugger that helps validate and troubleshoot expressions using simulated and authentic production data.
FogHorn’s platform also supports automatic edge device registration, one-click multi-edge deployment, multi-tenancy, and edge-based health metrics and alerts. It’s able to run models such as Spark ML and R Studio through the use of predictive model markup language, as well as computer vision models in a range of frameworks trained in the cloud and pushed to local devices like Raspberry Pi systems, internet of things gateways, and programmable logic controllers.
FogHorn says its platform and field service tools for Android have been used to develop bar code scanners, device health and battery monitoring apps, portable factory environmental monitors, and smart power tools. “Industrial IoT is at a tipping point. We’re seeing initiatives move quickly into full-scale deployments, especially among industries we serve in Korea and throughout the globe,” said LS Corp. chief strategy officer Yumi Lee, a FogHorn investor, in a statement. “FogHorn has pioneered edge analytics, [machine learning], and recently edge AI since its inception. LS looks forward to supporting the company as it continues to enable edge AI deployments and digital transformation initiatives across global industries.”
LS Corp. led the series B round, with Forte Ventures participating and existing FogHorn investors Dell, Intel, Saudi Aramco, Honeywell, GE, Bosch, March Capital Partners, and Darling Ventures contributing. Coinciding with the investment, John Neville and Senthil Kumar joined the company as chief revenue officer and VP of engineering, respectively.
Author: Kyle Wiggers.
Source: Venturebeat
