Falkonry's Machine-Learning Edge System Needs No Data Scientists

Falkonry is taking its predictive analytics tools for industrial companies to the edge, which it said is a first for a machine-learning system that doesn't require data scientists.

The Sunnyvale, Calif.-based industrial Internet of Things vendor said that its new Edge Analyzer system, which was unveiled Tuesday, is a "portable self-contained engine" that can run predictive analysis in edge devices for low-latency or offline environments.

[Related: Wind River Future-Proofs Industrial IoT With New Helix Edge Platform]

Edge Analyzer is an expansion of its Falkonry LRS machine-learning system, which already has capabilities for cloud and on-premises. Falkonry aims to help industrial companies improve operations, throughput, quality and safety while also reducing downtime by using an unsupervised machine-learning technique that identifies patterns from time-series data.

With Edge Analyzer, operators can now install Falkonry's predictive analytics application in remote or mobile environments with low hardware requirements. The application is containerized, meaning that it can be insulated from other processing activities. It also has tolerance for sensor and network outages and can be used to monitor multiple endpoints.

Sanket Amberkar, Falkonry's senior vice president of marketing, told CRN that the company is selling Edge Analyzer with a perpetual use license, which means Falkonry plans to make money with the hope of expanding deployments with customers over time while they pay a subscription for the system's cloud and on-premises capabilities.

Amberkar said one of Falkonry's largest advantages in the sea of machine-learning and predictive analytics tools out there is that it can be used by operators in the field.

"We're actually the first machine-learning company that has a product that can be used by non-data scientists to enable them to do the operations all the way from cloud to on-premises to the edge, and that's pretty unique,” he said.