Graph-based digital twins to underpin Australia's future sustainability reporting

With mandatory and “high-ambition” ESG reporting requirements on the horizon, organisations need a way to collect big data and bring it together.

Australian organisations face mandatory sustainability reporting requirements from the start of next year, pending passage of related legislation. As is the case in other geographies, such as the EU, extra time is being afforded to smaller organisations.

The government’s most recent guidance suggests that climate disclosure standards will be drafted by the Australian Accounting Standards Board (AASB) as early as August, “prior to the commencement of reporting requirements”.

But the government has also previously positioned its rules as a bare minimum — with a clear expectation that organisations do more with sustainability data than only what’s necessary. It wants to “encourage high-ambition approaches by firms, institutions and entities, including where these go beyond baseline standards established by regulation and policy”.

The policy shift in Australia reflects a global increase in demand for environmental, social and governance (ESG) metrics and reporting.

The change isn’t anything new for listed companies, which must already produce sustainability reports “as a condition for access to capital”. But the 2300-odd ASX-listed companies represent a small fraction of Australian organisations to which mandatory rules will apply.

Some Australian organisations have also previously reported voluntarily. These efforts are commendable but lack standardisation in formatting and disclosures, so will likely have to change to meet more stringent requirements.

Any change in ESG reporting and disclosure requirements on organisations will be felt acutely at the data level. It is necessary to collect and analyse a large amount of data to populate and produce sustainability reporting, yet many current methods are unlikely to scale to meet the increased reporting requirements that Australian organisations face.

Collecting and modelling sustainability data has traditionally been time-consuming and complex. For organisations, it’s not just about measuring their own direct CO₂ emissions, but also those generated in the upstream and downstream supply chain. Not every organisation in that chain has the same level of capability or data integrity to be able to do that accurately. Some of the estimates produced as part of voluntary efforts are imprecise and unlikely to meet audit standards.

Some industries, such as banking and finance, face more specific types of scrutiny than others around what businesses they invest in, and how that investment profile exposes them to emissions-related risks.

Given uncertainty about when and in what form enabling legislation may pass parliament, there may be a temptation for organisations to put off ESG data-related works until later in the year. However, this approach may place the organisation in an unnecessary time crunch. Since it’s not a case of if but when for mandatory ESG reporting requirements to be imposed, it makes sense to take action today to set up data reporting and modelling capabilities to meet tomorrow’s reporting and corporate responsibility requirements.

Managing a revamped ESG reporting process

Leading organisations in the space are looking to a combination of graph database technology and a digital twin to manage the complexity of the entire ESG reporting process.

In recent years, databases based on graph technology have established themselves as the solution when it comes to storing and retrieving vast amounts of data and analysing their complex relationships. This is mainly due to the fact that graph databases completely link the stored data with each other. They consider the relationships between data points (edges) to be just as important as the data points themselves (nodes).

To understand how this works, think of a graph database as a metro map: the nodes are the stops and the edges represent the tracks between stops. An algorithm runs through these nodes and edges, just like a train travelling from stop to stop. This allows the graph to understand what connects entities.

The system of edges and nodes makes graph databases transparent, dynamic and almost infinitely scalable. New data can be added and queried in real time. This allows organisations to realistically model large volumes of heterogeneous data. Simple knowledge databases with flat structures and static content can hardly achieve this.

In order to utilise graph technology comprehensively for ESG reporting, it’s worth using it as a framework for a digital twin. This allows organisations to model huge real systems in real time and map reality as accurately as possible. In this way, many different parameters such as CO₂ emissions, workforce diversity or investments can be monitored.

A graph-based digital twin can also be used to map the entire value chain with all its data. This results in a fully contextualised view of each unit in the value chain. With just a few clicks, it is therefore possible to recognise the extent to which unit X contributes to emissions, whether regulations are being complied with or whether there are critical dependencies. Paths with the lowest CO₂ emissions, potential sources of error or the causes of excessive emissions can also be identified.

In practice, the combination of graph technology and digital twin can be used to analyse both direct and indirect emissions. Furthermore, the technology provides the ideal basis for simulations of various ESG-relevant scenarios and for a semantics-based recommendation system for ESG documents.

All of this can save businesses valuable time, helping to ensure compliance with all legal and high-ambition ESG requirements, securing an enterprise’s own future and, more importantly, the planet too.

Peter Philipp, General Manager – ANZ at Neo4j.

Top image caption: iStock.com/Eoneren