Low voltage electricity networks are undergoing significant change as distributed generation connects to the grid and loads are electrified. At the same time, the number of sensors in low voltage systems is increasing through the installation of new monitoring systems (such as PMU's), and new measuring systems (such as smart meters). The growth in sensors means a growth in data that has the potential to be used in a way that benefits system operators, utilities and consumer. However, access to much of the data is very strictly controlled by privacy and security regulations, meaning the potential value provided by the data cannot be realised. 
Recent advances in federated machine learning techniques may provide a potential solution. This talk will investigate areas where federated approaches are being applied to smart meter data to help distribution system operators adapt to decarbonisation. The talk will introduce some of the opportunities for federated learning in power systems, describe the potential benefits of using federated learning when accessing private consumer data, and explain how the ongoing SFOE KnowlEDGE project harnesses them.