WASHINGTON - Employees at the National Institutes of Health are expected to set their computers on standby when they leave work at the end of each day. If they don't, a piece of software automatically does it for them.
IT officials said the policy saves the agency $19 per computer per year. That may seem like a sliver of the budget, but when multiplied by more than 25,000 terminals, it puts a noticeable dent in expenses.
NIH is among a growing number of government agencies, companies and organizations trying to reduce energy expenses by using information technology - hardware, software, big data, analytics - to better manage their buildings and rein in consumption.
IDC Energy Insights, a market research firm, projects spending on smart-building technologies to grow from $5.5 billion in 2012 to $18.1 billion in 2017, driven largely by a desire among companies to be more sustainable and cost-efficient.
While these systems may improve an organization's environmental footprint and budget, they also come at an expense - the costs of installation, materials and software licenses, as well as the costs that come with changing long-entrenched habits.
At NIH, for example, employees' penchant for always-on computers presented the biggest barrier to implementing the new software called Verdiem Surveyor, officials said. Plus, most of the agency's 27 institutes have independent IT departments that must accommodate the unique needs of different scientific researchers and physicians.
"There's a high regard here for [the idea that] any dollar that can be saved is a dollar that can be put toward research," said Daniel Lonnerdal, the agency's acting chief information officer in the Office of Research Services and Office of Research Facilities. "Even if [changes] may seem daunting or difficult at first, the longer picture is what's most important for most people."
The software is one of many measures NIH has implemented to curtail its energy use, including the purchase of light bulbs, computer monitors and other devices that are built to gobble less electricity, said Greg Leifer, head of the energy management team in the Office of Research Facilities.
Software to manage energy consumption in commercial buildings has grown more sophisticated in recent years as individual lights, computers or air conditioners can be connected to and controlled from a larger system.
"All of that instrumentation obviously generates a huge volume of data, so you have software that sits over top that utilizes that information so you make decisions based on it," said Casey Talon, a senior research analyst at IDC Energy Insights.
"The challenge where we're at is there are very sophisticated solutions and it's really a matter of seeing the business case for these solutions," Talon added.
Companies such as Arlington, Va.,-based GridPoint already are making the sales pitch. The company can outfit commercial buildings and facilities with sensors and observe energy consumption patterns remotely.
The firm's software has proved especially popular in industries that have a large number of isolated locations such as grocery stores, chain restaurants, banks and convenience shops, said Mark Danzenbaker, vice president of product management.
"We'll typically start out with a pilot of a handful of sites and really boil this down to proving the value. . . . Then we transition to a larger deployment," Danzenbaker said.
Once all the locations are plugged into the network, property managers can view energy consumption in every location from one program and assess ways to reduce costs, such as regulating thermostats and lights or identifying faulty equipment.
"It's all about the analytics. That's where the value actually is," said Kim Bishop, vice president of data services and business intelligence. "It's about the analysis that you put on top of [the data] and how easy you make it for your customers to consume those analyses so they can quickly take action, quickly save money."
Or, as is the case at NIH, officials can program the software to do the work, reducing the need for people to be involved in decision-making on such a detailed level.
"As systems and tools get even more developed and more reliable, the more they can be automated the better," Lonnerdal said. "It cuts down on not only the energy cost but the resource cost of having to do it manually."