Software startup brings predictive analytics to fire detection

MOUNT HOREB, Wis.—OneEvent, a software-as-a-service company based here, developed an analytics engine that looks at environmental factors to predict when an event might occur. At ISC West 2017, the company will be discussing the software's capabilities in the fire detection space.

“We look to the fire and security industries as a very natural launching point for OneEvent's software solutions and technology,” Kurt Wedig, OneEvent Technologies president and CEO, told Security Systems News in an email interview. “In UL research testing, the OneEvent analytical engine anticipated fire up to 20 minutes before the smoke alarms went off and in an event like that, 20 extra minutes can prevent a lot of damage and even save lives.”

OneEvent connects sensors into its cloud hosted predictive analytics engine, also named OneEvent. The company is now going to market with its first product containing the OneEvent software, OnePrevent. “We take various measurements within the environment, whether it be temperature, light motion, intrusion, smoke, CO, water. All of those different factors are put into the analytics engine and a predictive notice is given, based on the data collected,” Avi Rosenthal, a member of OneEvent's advisory board and a spokesperson for the company, told SSN.

The engine works with both OneEvent sensors as well as sensors from other manufacturers.

“The data analytics engine has the ability to look at many, many different things. The way that it's built is it's not specific to an individual sensor, it's specific to data coming in and data going out,” said Rosenthal. The system has applications across many verticals, residential and commercial, he added—any environment where there are “normal” and “abnormal” conditions.

Wedig said that early feedback on the system has been positive. “The use-case applications are so broad and the system can cover so many different aspects, from fire, to motion, to water, etc., that dealers get really excited by the flexibility of the system.”

The company will use a dealer model for distribution and will be looking for new dealers at ISC West.

“Over the next year, we will be very focused on expanding our dealer network,” Wedig said. “We offer security dealers and integrators a truly unique opportunity to provide a new, innovative service that is not available anywhere else. Low barriers of entry and ease of installation enable quick sales of the OnePrevent system for dealers and a clear path to increasing their RMR.”

Rosenthal said he expects other manufacturers will be interested in incorporating the OneEvent engine into their products, “But, like most startups in the IoT space, [OneEvent has] to prove themselves.”

The system learns from its environment, particularly with what is—and is not—a concerning event. “The more systems that are out there, the more data that we collect, the better we are at discerning the difference between a false positive and a true positive,” Rosenthal said. The company is also able to aggregate data anonymously across users to better learn the signature of a given incident.

OneEvent alerts are sent directly to an alarm user. Users can interact with a connected app to give feedback on an alarm and help the system learn the “signature” of an event, which will help the system distinguish between a threat and a less worrying scenario. “The more events that we see in your environment, the more signatures we have. The more signatures we have, the better we are at discerning when it's a true problem versus [when] you're burning the toast,” Rosenthal said.

OneEvent is working with insurance companies concerned with risk mitigation. “We're working now with a number of [insurance companies] to show them how predictive analytics can help them mitigate that risk,” Rosenthal said.

“We look forward to educating not only the security industry but others as well including fire and life safety, reclamation, and property/casualty insurance about deep learning and just how powerful predictive analytics in the future,” Wedig said.