In February 2018, researchers from the University of Queensland (UQ) School of Information Technology and Electrical Engineering announced the development of a world-first data platform that would harness information from potentially millions of households to enable deep, real-time analysis and control of household energy. The big data collected would not only empower consumers to take control over their household power generation, storage and consumption, but would also enable monitoring of entire systems and could be used by energy retailers to ensure safer and more secure networks.
According to Professor Neil Horrocks, Director, Centre for Energy Data Innovation at UQ, the platform and associated data use cases are still under development and at varying stages of completion, but the entire platform should be fully operational by 2021.
“The data platform itself is now in its second iteration, having been built and tested, and is now accepting data. It continues to evolve as new improvement opportunities are identified,” Mr Horrocks said.
“Data is being received into the platform every minute in near real time from sensors across the Energy Queensland network. Currently, we are receiving about four million sets of data values a week and this is growing by the month. The associated use cases that leverage the platform data are at varying stages of development.”
The university is working closely with four collaborators on the project:
- Energy Queensland Limited and specifically, its networks, Energex and Ergon Energy. The utility is overseeing the development and iteration of the use cases, installation of the IoT sensors and is providing additional existing network data flows
- Redback Technologies is providing the IoT sensor hardware for the project and the software engineers that maintain the platform
- Springfield City Group is providing ‘living labs’ in the form of energy consumers and its contribution will continue to grow throughout the project – as the developer of Australia’s largest masterplanned community, the business has an active interest in delivering innovative energy outcomes for its residents
- Microsoft is ensuring that the project team has the skills to deliver a reliable, secure and scalable platform product
“Most of the data is collected using IoT sensors that capture the data in near real time and transport it back to our Microsoft Azure Platform on a LPWAN. The sensors can collect a range of power data metrics and encrypt the data for transport,” Mr Horrocks said.
“The cloud-based Microsoft Azure platform has been customised by the software engineers at Redback Technologies to meet the specific needs of the project partners and ensure all the data is received and individually identified, validated and stored.”
Data with a purpose
Mr Horrocks said that two of the big challenges in the emerging big data world are obtaining access to the skills required to analyse the data and avoiding costly collection without clear use cases for that data.
“The Centre for Energy Data Innovation has a very clear mandate to deliver outcomes across a range of use cases,” Mr Horrocks said.
“It has assembled a multi-disciplinary research group to deliver this, including:
- Power engineers who provide context behind the use cases we are exploring and work with our data scientists to identify new learnings
- Data scientists who undertake the detailed analytics and machine learning design, and then constantly iterate their work to improve results
- Interaction designers whose role it is to visualise the outputs so they can be understood more readily by those using the data
- The project team also relies on the UQ cyber security team
“In addition, the team have a very clear project focus and have been working on a range of very specific data challenges, including:
- Working with households to help them to understand their power usage. Prototype visualisations are being developed to enable them to better understand their energy use
- Improving household safety by exploring a number of ways to detect the emergence of broken neutrals at a supply point using near real-time data (voltage, current, harmonics) to increase detection frequency
- Building improved visibility of connection points within feeder segments to assist networks to understand the network connectivity model and feeder capacities
- Tracking network power quality and reliability in near real time
- Forecasting short-term (day ahead) and medium-term (month ahead) solar generation profiles using detailed weather forecast data
- Exploring how IoT devices could provide better management of intermittent demand constraints and more targeted and timely demand response in near real time
- Exploring visualisation scenarios for network engineers to enable them to absorb more network data in a more effective and timely manner
“One of the features of the platform is that, using machine learning, it can make predictions and decisions about how consumers can most effectively use their energy.
“The project team are currently developing this capability and working closely with households to understand and uncover their energy needs, and to prototype options that will enable consumers to take charge in the future.
This is an iterative, collaborative process and there is still much to do. We expect the work will start to bear fruit in 2020 as we start to derive better insights into energy usage and develop improved visualisations and communications methods.”
Gaining consumer trust
With effective data security and privacy key to the success of such a platform, Mr Horrocks said that the project benefits greatly from having Energy Queensland integrated in the project team.
“They retain a high level of trust with customers developed through years of providing a reliable service and managing customer meter data. They are responsible for enrolling participants and providing security and privacy assurances.
Understandably, this in turn makes security and privacy a key issue for the project and a prime focus area for the project team,” Mr Horrocks said.
“The project team is closely watching the development of the Consumer Data Rights legislation and the mechanisms that will provide to allow consumers to safely arrange transfer of this energy data to their trusted parties.
“There are no plans to release the data externally, but work will be done to explore how this data could be effectively de-identified so that independent businesses might have access to representative data sets for product development and product testing.”
When asked if the platform would be made available to energy providers in other states, Mr Horrocks said that the design of the platform can be readily duplicated to allow other energy providers to have their own big data platform.
“This makes the platform an ideal way for energy providers to develop some early big data insights and test new opportunities without the cost or risk of building enterprise-sized big data infrastructure,” Mr Horrocks said.
“There are a number of networks in other states currently trialling the platform and sensor technologies with our project team.”
Interested in learning more about the threats to Operational Technology and how your organisation can guard itself against them? Click here to download the Report on Cyber Threats to Operational Technologies in the Energy Sector.