Smart home technology may soon help communities, and possibly entire countries, become more energy efficient. By designing an energy management system (EMS) able to collect and process big data into useful analytics, researchers from the American University of Sharjah, aim to help smart home owners and utilities reduce energy consumption, while improving the effectiveness of the power grid.

An EMS helps electric utility grid operators monitor and optimize energy transmission. However, current systems are unable to collect enough data to create energy use comparisons between a large number of households. Furthermore, smart home technologies are not interoperable with today’s EMS. As more homeowners invest in devices that are network-connected to control functions like lighting, temperature, security and entertainment systems, energy management systems have an opportunity to gather much more data about a home’s energy use.

“Most energy management platforms lack an efficient architecture to effectively manage a large volume of data from smart homes,” said Imran A. Zualkernan, PhD, an associate professor at the American University of Sharjah. “We developed a system with a state-of-the-art architecture that could handle data from users on a community, state and country level.”

Using the new EMS, each home device is interfaced with an IoT sensor and a unique IP address, which creates a large mesh wireless network of devices. A system on a chip (SoC) module collects energy usage data from the network and then transmits the data to a centralized server for further processing and analysis.

Figure 1: System Architecture

To transmit sensor data from the smart home devices into the centralized EMS, the researchers developed a message communication protocol using Message Queuing Telemetry Transport (MQTT). MQTT is able to transport data from a large number of devices because it is lightweight, meaning it transmits only functional sensor data to limit overhead as much as possible. It also improves device interoperability, which leads to scalable systems.

Along with the communication protocol, the researchers implemented storage, an analytics engine and web servers in the system. The storage server runs on the open source platform Hadoop to be scalable enough to handle storing vast amounts of data. To classify and process the data, the researchers used the Pentaho business intelligence tool and algorithms in its analytics engine server.

Once the EMS has collected data from smart homes in a community or region, the data can then be monitored and analyzed in real-time on a web application. Individual homeowners can access charts and graphs to help them increase energy efficiency, while utilities or regional authorities can access information to help them compare individual use to baselines. Homeowners also can monitor devices and access monthly bills.

Figure 2

Figure 2: Example of the EMS client application visuals provided for different user levels

To assess the scalability and speed of their EMS, the researchers created a prototype designed to mimic a small residential neighborhood of HVAC systems. The test showed the system’s MQTT protocol was able to process consecutive data transmissions for up to 10,000 clients. Its web server was able to process all requests for up to 4,000 users. Additionally, the simulations showed the storage server speed was also maintained for up to 4,000 users.

To further develop the system’s efficiency, the researchers plan to incorporate machine learning to mine and analyze data for more accurate predictions. They are also working to make it compatible with BACnet, which is a single building management system for HVAC systems. This would make the EMS interoperable with different HVAC systems in a smart home or building.

While technologies exist today to collect data from home sensors, managing that data and extracting insights from it remains a challenge. Using the new EMS, smart homes – and the big data they generate – may soon create new opportunities for smarter energy management across communities, regions and possibly entire countries.

To learn more about energy management systems, visit the IEEE Xplore Digital Library.