Abstract:
One of the most efficient ways of reducing household energy consumption, from a software engineering perspective, is through providing feedback to users on their energy consumption (energy feedback). Energy saving recommendations, visualisation of energy consumption and control of the appliances are some of the widely considered means to help end-users to conserve energy. Firstly, the existing literature and commercial solutions provide recommendations based on either usage patterns or context-aware data, but do not take into account user constraints and user preferences, and are limited only to basic recommendations. Secondly, different research solutions on visualisation provide aggregated & disaggregated energy consumption, near-real-time power data, and energy prediction, but not the location of the energy consuming appliances along with the energy consumption and appliance status. Moreover, the recommendations are limited to notifications and are not rendered within the visualisation for effective and immediate decision making by the end-users. Thirdly, the current research on control does not include controlling the appliances based on the recommendations and user preferences, such as timing constraints and targeted energy consumption. Finally, there exists no single system which integrates these three aspects; recommendation, visualisation, and control. We have designed and implemented a web application called ECoS (Energy Control System), a simulation model using a combination of user-centred technologies, such as survey and design thinking activity. The features of ECoS are: (i) recommendation engine - that provides recommendations to the end-users based on energy usage patterns, context-aware data, users' energy constraints, and user preferences;(ii) visualisation engine - that visualises household electricity consumption with exact location of where the energy is being consumed, along with the recommendations and control features; and (iii) control engine - that enables users to control and schedule their household appliances based on the recommendations and user settings through visualisation. The ECoS integrates the visualisation, recommendation, and control engines, and is evaluated by usability testing and quantitative experiments. An average of 21.61% household energy savings has been observed when tested on datasets of four countries. As a result, ECoS comprises of an optimal set of features based on user preferences, energy savings, and frequency of feedback, and it is hoped that ECoS would serve as a promising application for saving household electricity.