Pervasive and Mobile Computing ( ) – Contents lists available at ScienceDirect Pervasive and Mobile Computing journal homepage: www.elsevier.com/locate/pmc User activity recognition for energy saving in smart homes Pietro Cottone, Salvatore Gaglio, Giuseppe Lo Re, Marco Ortolani ∗ DICGIM, University of Palermo, Viale delle Scienze, ed. 6 - 90128 Palermo, Italy article info Article history: Available online xxxx Keywords: Activity discovery Peak load avoidance Structural modeling abstract Energy demand in typical home environments accounts for a significant fraction of the overall consumption in industrialized countries. In such context, the heterogeneity of the involved devices, and the non negligible influence of the human factor make the optimiza- tion of energy use a challenging task; effective automated approaches must take into ac- count basic information about users, such as the prediction of their course of actions. Our proposal consists in learning customized structural models for common user activ- ities for predicting the trend of energy consumption; the approach aims to lower energy demand in the proximity of predicted peak loads so as to keep the overall consumption within a predefined range, thus minimizing the impact on the end users. In order to build the models, the inherent recursive structure of user activities is abstracted from raw sensor readings, via an approach based on information theory. Experimental assessment based on publicly available datasets and synthesized consumption models is provided to show the effectiveness of our proposal. © 2014 Elsevier B.V. All rights reserved. 1. Introduction The ever-increasing energy demand in recent years is becoming a major issue as it represents a possible drawback in our society’s future development, where energy is arguably the single most valuable good. Current consumption trends are unsustainable from an environmental point of view, and efficient usage and overall energy demand reduction have become two major concerns of the international community and most governments, due to both economic and environmental motivations [1]. Namely, according to the classical market laws, those trends have caused a burst in energy price which eventually has attracted greater attention to the energy problem. The periodical shortages in energy supply during the last century, led to the birth of new research areas, and considerable effort is being carried out to devise viable solutions to the energy issue, ranging from discovering new energy sources to raising people awareness. In this context, a steady attention has been devoted to energy saving in buildings, starting from the energy crises of the 1970s [2,3]. User habits play a central role in household energy demand: an inefficient control of electric appliance and heating systems is a major energy waste source. Current literature about building automation, however, shows that building control is still mainly performed manually, as in the case of artificial lighting setting, powering appliances, or seasonal control of heating systems; additionally, automation in buildings has historically focused on narrow-scope tasks, such as lighting control with simple motion detection and a fixed timeout, or indoor climate control based on temperature and CO 2 level. On the other hand, user activities and behavior have considerable impact on the amount of consumed energy in all kinds of ∗ Corresponding author. Tel.: +39 09123862606. E-mail addresses: pietro.cottone@unipa.it (P. Cottone), salvatore.gaglio@unipa.it (S. Gaglio), giuseppe.lore@unipa.it (G. Lo Re), marco.ortolani@unipa.it (M. Ortolani). http://dx.doi.org/10.1016/j.pmcj.2014.08.006 1574-1192/© 2014 Elsevier B.V. All rights reserved.