Some residents of a small town in Iowa have cut their electricity use by 11 percent in a test that combines smart meters with cloud-based data analytics.
Utility company Alliant Energy installed smart meters at 1,000 homes and businesses in the town of Dubuque, which is close to the borders of Wisconsin and Illinois and has a population of about 60,000.
The smart meters report energy consumption to the utility every 15 minutes, and at the end of each day the data is shipped off to an IBM data center 600 miles away in Lexington, Kentucky, for analysis.
Residents can then log into an online portal to see information such as how much energy they have used that month, how it compares to previous months and how it compares to other households with similar characteristics.
Millions of smart meters have been installed in the U.S. but they’re being used mostly as a more convenient way for utilities to bill their customers, said Mark Bramfitt, an independent energy industry consultant.
It’s hoped that smart meters will also help consumers reduce their energy use, but engagement levels have sometimes been low. Just recently, Microsoft and Google both scrapped their online home-energy-management services, citing a lack of adoption.
One of the goals of the Dubuque test is to show that getting consumers more engaged, by having them compete against each other or use social networking to exchange tips, can allow such projects to succeed.
“Our theory is that without engaging the end user, we will not see the kind of projected results that we’ve been told smart meters will unleash,” said Milind Naphade, program director for IBM’s Smarter City Services Research.
The initial results in Dubuque are promising. Of the 1,000 households with smart meters, 800 agreed to take part in the test, and 200 of those have logged into the website to track their energy use. The industrywide average for “energy portal engagement” is 8 percent, Naphade said, so the test in Dubuque has exceeded that.
Sixty households have gone a step further and signed up for “activity groups” built around particular projects, such as installing energy-efficient light bulbs. The households in activity groups have cut their energy use by 11 percent, Naphade said. The test will run from July to December so the results are preliminary.
IBM’s software analyzes the meter data and puts households with similar characteristics into groups. They are then told where they rank compared to their peers. “We tell you whether you are number 15 or number 30 or whatever. We’ve found that to be a significant motivating factor,” Naphade said.
The smart meters provide only basic consumption data, but IBM said its Cognos software and other analytics products can figure out how much is consumed by big or small energy loads, and to “vampire” appliances that suck up power when not in use. It can also spot anomalies in a particular household.
“If we notice your air conditioner accounts for 45 percent of your load in the summer, and we’ve noticed other people with similar backgrounds consume only 30 percent, we can say, ‘Would you consider setting your thermostat 2 degrees higher?'” Naphade said.
If the customer agrees, IBM asks if it can track their consumption more closely to see what impact the change in thermostat temperature actually had.
“That’s the key piece,” he said. “You can go to all kinds of energy portal sites, but there are none that actually connect actions with their impact.”
IBM said the data is collected anonymously: Users are assigned a number that lets them view their own usage stats, but IBM sees only the number and can’t match it to an individual household.
“Smarter cities are the wave of the future when it comes to sustainability,” Dubuque Mayor Roy Buol said in an interview. Providing real-time data about energy use allows businesses and individuals to save both money and natural resources, he said.
If the test is successful, the eventual goal is to deploy it citywide, Buol said. Dubuque and IBM conducted a similar test with smart water meters last year that reduced water consumption by 6.6 percent, he said.
Covering the costs of such projects still needs to be worked out. “The utility industry will figure out how this will get paid for,” Naphade said.
Still, it remains to be seen how engaged consumers are willing to be in managing their energy use, Bramfitt said. They may prefer services that operate in the background, switching an air conditioner on automatically when a person is heading home from work, for example.
“People want to log into Facebook,” Bramfitt said. “But it’s not clear yet if they want to log into an energy portal.”
James Niccolai covers data centers and general technology news for IDG News Service. Follow James on Twitter at @jniccolai. James’s e-mail address is firstname.lastname@example.org