Almost a year after altering its business model, SHIFT Energy of Saint John is on track to complete two pilot projects that should validate its automated energy optimization software, or EOS.

SHIFT Energy is one of several industrial Internet companies in New Brunswick, meaning the province is developing a group of startups developing machine-to-machine learning software in which equipment reacts to disparate sensor readings to improve efficiency or save money.

The company was spun out of Saint John-based Mariner, a global provider of IP video monitoring technologies and other information and communication software. It was incorporated in 2009 to develop data-based products on energy consumption of large facilities. The idea was to help companies understand where and how they consume energy in their facilities so they could take action to reduce consumption.

In May 2013, the company changed direction. Instead of providing just energy monitoring software, it started building EOS — a solution that would react automatically to the data it collected. With the software installed, a large building or group of buildings can automatically reduce energy use in key areas based on data analytics. To achieve this goal, EOS uses a technique called Intelligent Live Recommissioning, which applies sophisticated algorithms to the data to determine the optimal settings instantly through the existing building control systems.

 “We used to do analytics and data collection and place it in the hands of the building’s operators,” said CEO Brock Sansom in an interview last week. “Utilizing existing building control systems already installed in a building, EOS can automatically manage the behaviour of the building on the operator’s behalf.”

Sansom said that during the company’s first few years, it gathered about 25 to 30 clients that were interested in data collection. Of these, two have opted to undertake pilot projects with the automated service of EOS. One is the Vancouver Canucks of the National Hockey League, who are testing the software at the Rogers Arena. The other is a large health network in Toronto, which is now piloting the software in one of its health facilities.

Both of these clients meet the two criteria the company has set for its client base: that they be large energy users, and that they have complex energy ecosystems. The idea is that there are so many complexities in managing the energy expenditure at large facilities that an automated system reacting to signals is more efficient even than a crew of humans monitoring the building around the clock.

A typical facility generates more than 5,000 unique variables each minute that can affect energy consumption, according to material provided by the company. Proactively analyzing and reacting to this volume of data is not manually possible, so the automated system opens up a new area of savings potential.

Sansom said the pilot projects are proceeding well and should be completed in the spring. The company is hoping to move into more live-event venues in the near future, and Sansom said its pipeline of prospective clients is growing.

The other members of the Industrial Internet Group in New Brunswick include RtTech of Moncton and Eigen Innovations, Xiplinx and SmartSkin Technologies, all of Fredericton.