Awareness on how and where energy is consumed is being increasingly recognized as the key to prevent waste in next-generation smart buildings. However, while several solutions exist to monitor energy consumption patterns for commercial and industrial users, energy reporting systems currently available to residential users require time-consuming and intrusive installation procedures, or are otherwise unable to provide device-level reports on energy consumption. To fill this gap, this paper discusses the design and performance evaluation of the Tiny Energy Accounting and Reporting System (TinyEARS), an energy monitoring system that generates device-level power consumption reports primarily based on the acoustic signatures of household appliances detected by wireless sensors. Experiments demonstrate that TinyEARS is able to report the power consumption of individual household appliances within a 10% error margin. (c) 2012 Elsevier B.V. All rights reserved.