Security data analytics today, for the most part, remain a combined effort between software and human analyses. But with new technological advancements, the software could be doing more work for the users.
Security data analytics today, for the most part, remain a combined effort between software and human analyses. But with new technological advancements, the software could be doing more work for the users.
Giant Gray, a Houston, Texas-based company, formerly known as BRS Labs, has brought to the market multi-sensor fusion technology that learns anomalous behavior and alerts to security or operational threats.
The company’s patented, adaptive and self-learning algorithms “see” anomalies hidden in vast amounts of data from sensors and bring it up as insights that operators, data analysts, security professionals and other end users can act upon.
At the heart of Giant Gray’s technology is its Gradient platform, with applications in industrial, physical security and surveillance, cyber security environments, and forensics. Speaking to asmag.com, Fred Palma, VP of Product Management at Giant Gray explained what the company does.
“If you look at the oil and gas industry, for instance, there are all kinds of different sensors that are all showing temperatures, measurements, and so on,” Palma said. “We process that data, learn the normal patterns of behavior and alert when something anomalous occurs.”
The operators in the oil and gas industry work with setting certain limits to the sensors. If the limits are reached, they are alerted. However, quite often, when the sensors reach the high and low limits, it’s already too late. What Gradient is able to do is learn the patterns of behavior and make alerts when something out of ordinary happens so that customers can make operational adjustments before the early warning signals grow into true incidents that impact safety, security, and business operations.
Giant Gray claims Gradient is scalable, easily integrated, and compatible with current monitoring systems, addressing limitations of existing machine learning analytics and monitoring technologies without requiring predefined rules, models, custom programming or data analytics expertise.
Gradient also stands apart from traditional machine learning analytics with its artificial cognitive neurolinguistics approach, meaning it creates a custom language to describe data from scratch and is not limited to baseline behavior models from the start.
Prepared for IoT and Big Data
The company says the Gradient platform is designed with the big data and Internet of Things (IoT) vision, supporting businesses and government agencies of all sizes across industries. It helps reduce the costs and simplify usage and maintenance by teaching itself from observing data and detecting the unexpected.
“What’s happening particularly when you start looking at the IoT and the big data, is that there are all kinds of sensors that are going online, from video, traffic signals to pretty much everything,” Palma continued. “So what are we going to do with all that data? What we have to do is some filtering, which is analytics. What we do is analytics of big data from all these sensors that are now available online.”