Algorithm Development for Smart Home Software: The Home Utility Management System

The supply of power and water provide the support upon which sustainable and stable economies and communities are built and maintained. The current centralized power and water distribution infrastructure comes with some potentially serious issues, however. These include vulnerability to natural disasters, cyber/physical terrorism, and climate change. For example, limited damage to centralized systems can significantly reduce resource supply to large portions of populations, and water shortages can be catastrophic in increasingly arid regions. These issues can potentially be solved with a move toward decentralized utilities and renewable energy. The Home Utility Management System (HUMS) home is a cost-effective realization of this need that can maintain a high quality of life. The home produces its own power through a residential wind turbine and solar panels mounted to the roof, with surplus power stored in a series of batteries. An on-site water tank collects precipitation for use in the home. The home itself and its appliances are energy and water efficient.

In this study we describe the HUMS software and its algorithms. The HUMS combines high-quality probabilistic forecasts with past utility use to provide resource conservation guidance to residents. The algorithms loosely follow this flow: 1) associate water and power use with specific home appliances and equipment that use these resources; 2) predict upcoming resource demand from past use data and probabilistic forecasts; 3) predict expected resource recharge from probabilistic forecasts. The software condenses information produced by the algorithms into useful and easy to understand guidance for the residents. A web browser application installed on the computer included with the home displays the guidance messages, resulting in a unique, two-way sociotechnical model to address the issues described above.

The residents’ interaction with the software and their perception of quality of life afforded by the HUMS home are key components to evaluating the success of the project. Pairs of residents will be selected to live in the HUMS home for four separate six-month periods. Residents will be asked to provide feedback on their experiences through journaling, interviews, debriefings, and focus groups. We will continue to develop the software alongside these evaluations, with the goal that HUMS can infiltrate the commercial housing market at an affordable price for mid-range buyers.