As smartphones and other mobile Internet devices have become ubiquitous, manufacturers continue to cram more and more features into each new device. Each generation of device is able to do more things faster. Through innovative technologies like 4G and processor improvements from Moore's Law, users can access web content much more quickly and run applications and games previously only usable on desktops and laptops.
Despite these huge gains in functionality, one significant problem still remains for mobile devices: battery life. Devices continue to use Lithium ion batteries, where improvements are much slower than corresponding improvements in other components like processors. [1] As such, though your next smartphone will be able to do everything much quicker, you'll also have less time to use it before it runs out of power. Of course, the ideal situation would be for battery technology to keep up, but assuming we don't find a silver bullet to achieve this any time soon, we'll need to come up with more clever strategies to keep devices running as long as we'd like them to.
Most devices, mobile or not, implement some form of energy-saving interface, such as dimming the screen or entering sleep mode. A lot of these techniques have been around for awhile, and one might wonder whether they are still reasonable for mobile devices. Researchers at HP looked at this question by surveying users about different energy-saving interfaces. They presented users with different experiences such as gradient screens (bright at the top, dim at the bottom), and inverted color screens (bright text on dark background) so see how users felt about these strategies for saving battery life. They found that users were generally happiest with the inverted color screen because it still allowed the text to be easily readable. They also found that for tasks like playing music, users much preferred an interface that saved power rather than the conventional, more colorful interface. [2] These results suggest that manufacturers and developers should think about power management as they design their applications. Some techniques, like inverting the colors of the screen or making the user interface dimmer, can dramatically increate battery life without significantly inconveniencing the user.
The key to interfaces that help you save power is knowing when they're necessary. Today, nearly all devices base this decision on the percentage of battery left in the device. However, this assumes that the user plans to drain the entire battery every time. In fact, as anyone with a smartphone would know, this isn't usually the case. How much and how often we recharge our battery depends more on where we are and when we expect to be able to recharge it again, rather than exactly how much battery is left. Researchers confirm this intuition: they found that most recharges happen when the battery still has a lot of charge, and recharge patterns are based more on location and context than on battery level. [3] This means that activating energy-saving features based only on battery level is the wrong approach. Instead, devices should try to predict how much energy will be needed until the next recharge. Even if the battery is mostly full, if the device realizes that you're 500 miles away from home, it shouldn't operate as though you'll be able to recharge it in the next 2 hours.
Using this idea, researchers at Microsoft created a system to predict how long a user will go before recharging the device again. Using information about which cell a phone was connected to, they compared where the device is currently to where it's been in the past. By designating certain locations as "recharge stations" (based, for example, on where the device spends most nights), they could then predict how long it will be before the device enters the next recharge station. [4] The combine this predictor with a predictor for how much the user will need the device, and as a result, they are able to give users timely warnings about when the battery might not last until the next recharge. Nonetheless, they found that they were much less accurate for users with more unpredictable lifestyles, and they propose that additional information about hte user (such as his or her schedule for the day) could help.
Given that many recharges occur when the battery is still mostly full, a natural question to ask is whether a device can effectively use the excess energy. For example, if we know that the device will be recharged soon, we might want to improve the user's experience by, say, increasing the screen brightness or performing some tasks in the background that speed up web browsing. Researchers at Amhearst, Rice, and USF built a system that does just that. Basing recharge predictions on time only (without location information), the system estimates how much energy is available for "non-critical" tasks and enhancements. They found that users experienced much brighter screens and many faster page loads after installing their program, without being required to charge their devices any more than usual. [3]
These are just a few examples of how effective software design can help users make the most of their mobile devices' batteries. Even software that potentially shortens battery life can be useful for battery management if it accurately determines when the charge won't be needed. Interestingly enough, most devices today continue to rely on battery level percentages and energy-saving modes that only activate when the battery is very low. Before throwing our hands up and wishing that battery technology could keep up with today's users, perhaps we should first worry about having our software keep up with them and more accurately respond to their usage habits and needs.
© Michael Chang. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] M. Geuss, "Why Your Smartphone Battery Sucks," PCWorld, 18 May 11.
[2] L. Bloom et al., "nvestigating the Relationship Between Battery Life and User Acceptance of Dynamic, Energy-Aware Interfaces on Handhelds," in Mobile Human-computer Interaction - Mobile HCI 2004, ed. by S. Brewster and M. Dunlop (Springer, 2004), p. 13.
[3] N. Banerjee et al., "Users and Batteries: Interactions and Adaptive Energy Management in Mobile Systems," in Proc. 9th International Conference on Ubiquitous Computing, UbiComp 2007, ed. by J. Krumm et al. (Springer, 2007), p, 217.
[4] N. Ravi et al., "Context-Aware Battery Management for Mobile Phones," Proc. Sixth Annual Intl. Conf. on Pervasive Computing and Communications, March 2008, p. 224.