Newsroom: Highlights

Bookmark and Share Email Print

12.18.19 - Ready, Set, Sleep



There are more than 2 billion personal computers in the world today, according to data from Forrester Research, each plugged in and ever ready to process and perform.

Although laptops are growing in market share, in many homes they are supplementing rather than replacing desktop computers. “This is important because desktop computers use − and waste − more energy than any other consumer electronics device,” says Joy Pixley, research director of the California Plug Load Research Center (CalPlug) at CALIT2 UC Irvine.

Desktop computers come with built-in settings that put the computer or the monitor into sleep mode after a certain amount of idle time − that is, when the user hasn’t touched the mouse or keyboard for a while. In sleep mode, the average desktop computer uses about 2 watts of electricity, compared to about 48 watts in active mode. The trouble is that too many computers have their sleep settings disabled, so they stay on all the time, using almost as much energy while idle as when they’re being actively used.

“Computers have perfectly good sleep settings if they are enabled, yet we have increasing evidence that they are not enabled. Something is going wrong,” says Pixley, a behavioral researcher.

She has been lead or co-lead on several large-scale studies of computer users’ behaviors sponsored by agencies including the California Energy Commission’s (CEC) Research and Development Program and the Electric Program Investment Charge (EPIC). Her past work was instrumental in informing the CEC’s appliance standards that took effect on Jan. 1, 2019. These standards apply to desktop computers, laptops, small-scale servers, workstations and monitors, and could potentially save more than $370 million annually − enough to power about 350,000 average California homes for one year.

In an initial survey, completed in 2014, Pixley asked more than 2,000 people about their home and work desktop and laptop computers. Later, her team conducted a second study that monitored more than 100 of the subjects’ office desktops. Comparing the results of the two studies held a surprise: in the first study, 78 percent said the sleep settings were enabled. “However, when we went and observed those same desktop computers, only 20 percent actually had sleep settings enabled,” she says.

For the third and most recent study, Pixley focused her attention on the two-thirds of users whose sleep settings were disabled but who thought the settings were enabled or didn’t know. “How could this group be encouraged to voluntarily enable sleep settings?” she asks.

Two common techniques are used to promote energy efficiency with computer settings. Computers can be shipped to consumers with the sleep settings already enabled. “This works great,” Pixley says, “up until the moment somebody disables them.” Another option used by businesses with centralized IT departments is to control the workers’ computer settings. But IT departments don’t always welcome the possible hassle and pushback from employees, and it doesn’t work for smaller companies that don’t have a centralized IT unit, or whose employees have home desktops, she adds.

Previous interventions have tried giving people feedback on the energy consumption of their computers (or more often, their whole office), which does seem to reduce energy consumption. However, this requires special metering, and prior studies haven’t established whether users are enabling sleep settings (an effect that lasts), or simply turning their computers off more often during the study.

The goal for Pixley’s research group was to create a no-cost software application that provides users with feedback on how much time their computers were idle, in the hope it would encourage them to enable their sleep settings.

CALIT2 researchers Sergio Gago Masague and Raquel Fallman helped Pixley design and program the application, PMUI (Power Management User Interface), based on behavioral theory and intervention research.

A field test of more than 400 desktop computers was conducted in UCI’s campus offices to test the effectiveness of PMUI and to collect data on users' behaviors toward power management.

The intervention included three visits from the researchers. On the first visit, PMUI was installed in observation mode only to collect baseline data. Only 14 percent of the computers had their sleep settings enabled at the beginning of the study.

After one month, the control group subjects were shown how to locate their computers’ regular sleep settings, while the PMUI interface was activated on the computers of the treatment subjects, and they were shown how to open it. No one in either group was urged to change their settings. Two months later, the researchers returned for a final visit to remove PMUI and the plug meters.

The results showed that PMUI changed users’ behaviors for the better. After being shown the PMUI app at the second research visit, the treatment group was much more likely to enable their sleep settings than the control group, who were only shown their regular sleep settings. The difference was even greater two months later. Those who had the PMUI app were significantly more likely to keep their sleep settings enabled, or to enable them by the end of the study.

The effects of enabled sleep settings were found to deliver an energy savings of 34.9%. “If you have a large office building, that’s going to add up pretty quickly,” says Pixley.

“Many people don’t know how their computers are set; providing users with feedback on how computer settings impact energy use is valuable and can result in reducing cost to California consumers,” says Bradley Meister, CEC project manager. “As California pursues a carbon-free future, it is important to ensure that we maximize energy efficiency, and this software program provides consumers with the tool to do so and produce savings beyond the current state standard.”

One unexpected finding revealed that two-thirds of computers with sleep settings enabled experienced problems with transitioning into sleep mode, including 42% that spend at least 25% of the day idle when they should be sleeping. Something is blocking the settings and keeping the computer in idle mode. Pixley considers that peripherals and perhaps other applications are overriding the sleep setting commands and preventing the computer from shifting into a more energy-efficient state. Many programs interrupt sleep settings, such as virus checks, backups or system updates, she says, “but your computer is supposed to go back to sleep” in order to conserve energy. Computer experts know sleep block is a problem, but the complaints are anecdotal so it’s difficult to say how large the problem is. Pixley has been unable to find data on this subject. “I’m always hesitant to say this, but as far as I can tell, this is the first data set to quantify the extent of the problem of sleep block.”

Results from the study provide strong evidence that feedback on computer idle time can convince users to change their sleep settings, saving a substantial amount of energy − something no prior study had addressed.  “The exciting thing is proving that a free software app can change behaviors and save so much energy, without needing to install extra equipment,” Pixley says. “Giving people feedback, encouragement and engagement works.”

- Sharon Henry