Smart meters have become increasingly ubiquitous. The devices which record information such as consumption of electric energy, voltage levels, current, and power factor are commonplace in households across the world. Indeed, By the end of 2022, there were over 124 million smart meters installed in 78% of U.S. households, according to data released in April.
Our client, a major utility with operations in the electric power sector, was deploying a network of smart meters that could transmit data over a digital network. While these smart meters were installed to monitor customers’ energy usage and bill them accurately, our client recognised that they could also provide valuable data.
As such, they were interested in developing applications that would fully realise the potential of the data generated by their network of smart meters, but they needed expert guidance to evaluate the wide range of possible applications. They asked us to determine what kinds of applications could be developed using data management, and what new insights could be derived by analysing their smart meter data.
We assembled a project team of four experts, all of whom were selected for their experience with the kinds of unstructured data that our client’s smart meter network would generate. Our experts helped us perform a comprehensive analysis of smart-meter-based applications and analysis technologies, covering both current solutions and technologies still in development.
Our team also investigated machine learning algorithms that would allow our client to analyse their smart meter data at scale and create a range of new applications. Some of the application areas we considered included developing new tools that could benefit “prosumer” customers, who also produce energy that they provide to the broader electric grid. We also investigated the development of tools to improve demand forecasts and help our client manage their grid more efficiently and effectively.
The team identified additional opportunities, such as creating consumer-facing applications which use their smart meter data, and determining the best network communications technologies to use for their smart meter network. We also identified challenges and risks that our client needed to monitor while developing and deploying their smart meter network, including potential cybersecurity threats.
Our team’s conclusions illustrated the strengths and weaknesses of each technology relative to our client’s specific goals. Our framing helped them to understand which technologies were most mature, and to consider those that could provide the greatest benefits when integrated with their smart meter equipment and networks. As a result, our client is pursuing the machine learning-based solutions we recommended, and they are now developing internal tools for each use case and each application that our team identified.