These days, the perfect Aussie home among the gum trees doesn’t just need a clothesline and a veranda – solar panels are a must-have feature.
Solar panels are a type of Distributed Energy Resource (DER). DERs, such as electric vehicles and solar batteries, are small-scale units of power supply which can provide additional supply or reduce demand on the grid.
Our modelling with Energy Networks Australia forecasts that up to 45 per cent of Australia’s electricity will be generated by consumer-owned DERs by 2050.
These technologies are advancing at a rapid pace and becoming increasingly affordable. What’s more, the International Energy Agency reports DERs are a main driver of the global transition to net zero emissions.
DER we go: introducing the Internet of Energy
DER technologies will play a significant role in addressing climate and energy security challenges by introducing clean, renewable energy at scale. They play an important part in the Australian Energy Market Operator’s Integrated Systems Plan, which lays out a step-change scenario for the transition away from coal-fired power.
But ensuring DERs reach their full potential will take a lot more than just installations.
For DERs to help us reach net zero, we need to understand, monitor and manage the supply and demand for energy at enormous scale. This could soon be a reality, thanks to advances in artificial intelligence (AI) and the speed and quality of wireless networks.
Surya Nepal is leading an international collaboration with researchers from Georgia Tech in the USA. Together, they’re preparing for what they’ve dubbed ‘The Internet of Energy’.
Optimising renewable energy resources
“In the Internet of Energy, all of our DERs can be connected and integrated for reliable energy distribution by leveraging advanced 5G-6G networks and AI technology,” Surya said.
Many stand to win from connecting DERs to form a network of smart devices.
Energy operators can optimise energy resources and maximise performance across the system. Governments can develop informed policies and regulation. This is needed to increase the spread of DER technologies in communities and to ensure equitable energy access and pricing. Customers can save money by optimising energy production and consumption.
All of these benefits should ideally work together to incentivise more renewable energy generation and help us on the road to net zero.
AI drives smart, clean, energy machines
DER-supplied electricity is currently dependent on several variables, like the weather. A solar panel can generate more power on a sunny day than on a rainy day. However, current systems don’t optimise production and storage.
Surya describes Australia’s current DER system as “install and pray”.
“At the moment, the visibility is limited to what you produce, but there is little control and smartness in the system. We need our DER systems to be smart to cope with flexible requirements. AI can bring that smartness to an integrated system,” he said.
The predictive power of AI could be used at scale to forecast impacts on DERs like weather. They could also identify periods of peak demand to help manage production and minimise waste.
“In a smart system, customers will have a secure digital platform enabling an individual consumer to decide when to become a producer and a consumer. It can automate decisions such as when to consume, store, or supply to the distribution network,” Surya said.
“The system can find an optimal solution for an individual and their needs, but also consider the stability of the whole system.”
An equitable, responsible Internet of Energy
Currently, the penetration of AI technologies in DERs is very low. But not for long. A2Z Market Research predicts there will be massive growth in AI-enabled DER platforms by 2029.
But it’s critical that the future AI-enabled Internet of Energy be equitable and responsible. We need to tread with caution and care.
Firstly, privacy is a central concern. Techniques to protect the privacy of data are critical. We’ll need them to manage and analyse the sensitive location, time and energy usage data needed to inform effective energy resource management.
The system also needs to be secured against external threats. For example, the physical components of the Internet of Energy like DERs need to be resilient to extreme weather conditions. The underlying digital infrastructure (software, algorithms and data) need to be robust and protected against cyber-attacks.
Finally, the training of AI models used to inform the management of the system needs to be inclusive and represent a broad range of users from different locations and populations. For instance, privileged neighbourhoods are more likely to have a large instalment of roof-top solar panels. Training models purely on that data would result in model bias and an unfair system.
Harnessing the power of AI with the Internet of Energy
Together with our colleagues at Georgia Tech in the US, we’re working to enable a responsible AI-powered Internet of Energy. This includes developing safe, secure and equitable AI frameworks, algorithms, and compliance methods.
Computer science expert Prof Ling Liu is the joint project leader from the Georgia Tech. She said the collaboration seeks to ensure everyone can reap the benefits of advances in AI to provide safe, reliable and sustainable energy services.
“This joint research project sets out to develop equitable AI technology to make the DER market and the Internet of Energy ecosystem available for everyone, and responsible AI technology to safeguard DER monitoring, management, and consumption through ensuring AI privacy, AI security, and AI fairness,” she said.
The project is funded by a collaboration between the US National Science Agency and CSIRO to drive ground-breaking research in responsible and ethical AI solutions to address global and societal challenges.
The collaboration was acknowledged by the Prime Minister Anthony Albanese and President Joe Biden in a joint statement on the next generation of innovation and partnership between the US and Australia, released during the PM's official visit to the US.
Dr. Aaron Quigley is the Science Director for CSIRO's Data61. He said Australia's collaboration with leading researchers from the United States would help progress responsible and scalable solutions to some of the world's most pressing global challenges.
"Our ongoing collaboration with the NSF is investing in cutting-edge AI and collaborative intelligence. These projects help explore critical infrastructure protection and resilience, advancing research into critical digital infrastructure to help create a more equitable, safe, secure and sustainable world in our global transition to net zero," he said.
Across CSIRO, we are also working on protection and resilience for Australia's critical infrastructure, and building Australia's next generation of energy systems through our developing Critical Infrastructure Protection and Resilience and Smart Energy Missions.