Towards Privacy in Monitored Shared Environments

For the purpose of energy efficiency, automation and consumption visualization, an increasing number of sensors like power meters, temperature sensors, and CO2 sensors, are installed in private households. Such a monitored household, the Energy Smart Home Lab (ESHL), was built on the campus of KIT for research on energy management systems [1]. In addition to ambient temperature and thermal flows, the lab records energy consumption data of 37 electric metering points to monitor each appliance. The recorded data can be visualized in the form of graphs as shown in Fig. 1. Monitored environments like the ESHL are usually shared by multiple residents whose actions and behavior are reflected in the data. By recording and analyzing the data, inferences about the actions and habits of residents of a household can be drawn. Molina-Markham [2] and Lisovich [3], for instance, demonstrated that sensitive information like occupancy, appliance usage, sleeping cycles, and cooking habits, can be inferred using only energy consumption data. On the one hand, revealing such sensitive information to the involved residents is very important to improve energy awareness. On the other hand, sensitive information about other residents can be abused for continuous surveillance.