Improving mobile user experience in low-battery situations through real-time energy consumption prediction

Abstract

Low-battery anxiety has emerged as a result of growing dependence on mobile devices, where the anxiety arises with the battery level running low. While battery life can be extended through power-efficient hardware and software optimization techniques, low-battery anxiety will still remain a phenomenon as long as mobile devices rely on batteries. In this paper, we investigate how an accurate real-time energy consumption prediction at the application-level can improve the user experience in low-battery situations. We present ALBATROS, a mobile system framework specifically tailored to predict the energy consumption of each mobile application and present the prediction in a user-friendly manner. Using ALBATROS, we conducted user studies to verify that energy consumption prediction with high accuracy can effectively alleviate low-battery anxiety by assisting users in planning their application usage based on the remaining battery life. We summarize requirements to mitigate users’ anxiety, guiding the design of future mobile system frameworks.