Five minutes with CSIRO's Melissa James

30 Sep 2024

This month we spend five minutes with Melissa James, Senior experimental scientist at the CSIRO to find out about RapidRate, a new artificial intelligence system that estimates energy efficiency for homes across Australia.

Melissa James is a data scientist, project leader, and communicator with more than eleven years of experience in residential building energy efficiency research. She delivers science, data, software, and communications that support the shift towards energy efficient housing in Australia. Melissa is currently exploring ways to use Nationwide House Energy Rating Scheme (NatHERS) data for research which contributes to reducing energy consumption in Australian households. Melissa is leading the development of RapidRate, a tool for quickly rating the energy efficiency of dwellings, developed using machine learning techniques over NatHERS certificate data.

1. What makes measuring the energy efficiency of an existing home challenging?

When new homes are assessed to demonstrate they meet the minimum standard of energy efficiency required for building approval, most of them use a NatHERS accredited tool. Details from the building plan are entered into the NatHERS tool; the heat flows in the dwelling are modelled (based on climate, building materials, orientation etc.); and the amount of heating and cooling energy required to keep the dwelling a comfortable temperature is calculated. This process takes a long time, requires a lot of input data, is relatively costly, and requires expertise.

In contrast, existing homes are not required to undergo energy efficiency assessments, and there is no incentive for homeowners to do so. Although they could choose to have their home assessed using a NatHERS tool, barriers such as high costs and the absence of building plans often deter them.

2. Updates to the National Construction Code have seen a surge of new homes rated above 7-stars with NatHERS, which is a big step forward. How will RapidRate help existing homes?

It is estimated that homes built before 1991, when building energy efficiency regulations were introduced, have an energy rating of just 2.5 stars (out of 10). Although regulation ensures newer homes meet minimum energy standards, many older homes perform pretty poorly. Unlike new homes, which undergo assessment for building approval, there is little incentive for existing homes to be assessed to understand how they perform, and how they might be improved. A full NatHERS assessment is relatively costly, needs a lot of data inputs, and requires expertise.
To address these challenges, CSIRO developed RapidRate using Artificial Intelligence, trained with NatHERS data. As a result, RapidRate outputs are aligned with NatHERS ratings, but require far fewer inputs. It is quick, efficient and can leverage existing property data - such as that held by real estate agents and property data companies - to estimate how energy efficient existing homes are.

3. How will RapidRate work with or complement existing rating tools like NatHERS?

RapidRate has the capacity to fill the data gap relating to the energy efficiency of existing homes. Data generated by NatHERS tools provides information about newer homes, but there is very little data about the energy efficiency of existing homes. By combining RapidRate with existing property data, this gap can be filled relatively quickly. RapidRate can also be used to explore upgrade options for individual dwellings and assess their impact on energy efficiency.
NatHERS tools use physics-based models, give more accurate assessments, and are the best choice for regulatory purposes. RapidRate is great where a quicker assessment requiring fewer inputs is needed.

4. How is CoreLogic collaborating with CSIRO to use RapidRate?

CSIRO has licensed the use of RapidRate to CoreLogic – a national company with over 40 years’ worth of property data. By combining CoreLogic data with CSIRO’s RapidRate artificial intelligence model, an estimate of the energy performance of existing homes is generated.

5. How will RapidRate be used by banks initially and how do you hope to see it used in the long term?

Banks can use the RapidRate output data to understand the energy efficiency of their property portfolios, help with financial emissions reporting, and better support their customers in the reduction of carbon footprints.
RapidRate output data will enable insights into the energy performance of the country’s existing 11 million residential properties. Understanding the energy efficiency of Australian homes is the first step towards making them perform better. Energy efficient houses are healthier and more comfortable, cost less to run, and result in less carbon emissions.

You can see Melissa James, GBCA’s Elham Monavari and other speakers discuss data-driven design in sustainable housing at the upcoming Sustainability Summit. Find out more about this session on the Sustainability Building Awards website.