Laptop energy-saving opportunities based on user behaviors

Abstract

In order to improve the energy efficiency of laptop use and design, it is important to gain knowledge of how laptops are being used in real-world settings. However, the current workload studies have focused on computer-usage patterns of business users, but not those of consumer users. As part of the u! Green project at ABC University, a pilot study was undertaken to develop and test software tools and methods for monitoring the laptop computer usage of 13 college students for a period of 1 month. This paper describes such tools and methods for monitoring and collecting laptop usage data. It also illustrates how the collected data can be analyzed to obtain high-impact energy usage patterns, which include long idle periods, excessive use of the AC power source, and high backlight levels. These identified high-impact usage patterns, if confirmed by our subsequent study, offer valuable user behavioral insights to advance more energy-efficient laptop use and design.

Are computers agents? Considering the implication of classifying computers as occupants on energy consumption and proximity-asutility equipment scheduling

“…Discussion on disaggregation of occupant-attributed disturbances has received increasing attention over the last few decades. The suggestion that some equipment are closer to being occupants than equipment in the conventional sense has in principle been covered previously in activity detection [4,8] where appliance-level consumption was considered significant to understanding user behaviours. Jones, Wei [4] discussing system processes and energy consumption implicitly suggest the impact of idiosyncratic personalities of the machines themselves.…”

“…The suggestion that some equipment are closer to being occupants than equipment in the conventional sense has in principle been covered previously in activity detection [4,8] where appliance-level consumption was considered significant to understanding user behaviours. Jones, Wei [4] discussing system processes and energy consumption implicitly suggest the impact of idiosyncratic personalities of the machines themselves. Personality seems further supported by Joy E, Suart A [9] who note agents are not exclusive to a single computing devices or locations and Gunay, O'Brien [10] who identify significant differences in the energy conscious behaviours of respondents when using differing computer types.…”

“…However, the latter observation is contrary to those of Wang and Ding [14] who observed 90% of users switching off during predictable periods of absence They also notice the appearance of more energy-awareness from laptop users with 25% of computer users reporting never turning off in contrast to only 5% of laptop users and 10% laptop users shutting down during intervals. Jones, Wei [4] monitored energy consumption to extrapolate and potential energy savings from computer users suggesting global computer energy consumption could be reduced by 17TWh with lower thresholds for standby mode, and 230 -68GWh through reduction LED backlight intensity, per annum. Though they acknowledge their sample size is too small to make definitive conclusions.…”

“…Ethical quandary of ad-blocking aside, these show not only is equipment-level user behaviour monitoring desirable, but the personality of the system itself is worth considering. The problem extends to residential buildings where it is expected equipment energy consumption will tripled by 2030 from 2009 levels [4]. Zhang and Bai [15]'s study of consumer goods ownership in Shangdong identified a 600% increase in computer ownership per hundred urban households between 2005 and 2013 indicating a drastic surge in electronic energy consumers.…”

“…Previous building physics research discussed what could be considered the impact of equipment personalities on energy demand with [4] identifying an established case for equipment similar to the theme of this article proclaiming "simple workloads" fail to reflect the complexity and variety of user or equipment behaviours. Jones et al effectively discuss the hardware personalities and identify idiosyncrasies in the form of mixed focus software activity, thought they do not consider the relationship between the agent and mediator, the computer.…”