NBN Co turns to AI to improve network management

NBN Co is turning to AI and automation as it strives to maintain and improve service quality and reliability.

The company’s executive general manager for technology transformation Jo Dutton said AI is a useful technology for improving network management thanks to its ability to make use of the terabytes of data that the NBN network generated each day.

“What people expect from us is a reliable network that really allows them to do what they want to do and when they want to do it, so they just want it to work,” Dutton said.

“And that's why we see assurance of the network moving from transactional processes to proactive ones.

“Where we really see benefit from AI is about our ability to drive more advanced insight from our data sets, and then be able to apply that to our process automation and our decision making.”

Dutton described the current process for fault reporting, which first involved the customer alerting their retail service provider (RSP) and which could take up to 72 hours to remediate.

“If we look at both the technology we have today, but also the emergence of more accessible AI technologies, as well as things like robotics, we see a future where we're able to mine our data, spot the issues before the customer experiences anything, and either remotely or robotically enable a fix,” Dutton said.

The key to this capability is the massive volume of data that was generated by the network today – up to 10 terabytes on any given day – which provided a rich source from which AI could extract useful insights.

“AI is a tool used by humans to achieve things we wouldn't be able to cognitively achieve otherwise, and if I apply that to NBN, it's about really gaining those insights from that massive amount of data,” Dutton said.

“That's one of the really interesting parts of this is it's the ability to make decisions based on data sets that we couldn't possibly comprehend ourselves.”

While Dutton said the vision for an AI-enabled network is compelling, she is also cognisant of the many considerations that would need to be evaluated through its development.

“While the technology is evolving rapidly, we are taking data privacy and cyber security really, really seriously,” Dutton said.

“We're more focused on establishing an ethical and sustainable framework, so that as we do accelerate with the adoption of the technology into our day-to-day network performance, we can have confidence that we're doing it in a secure way.”

For this reason, NBN Co had established a framework within which it could conduct safe experimentation to better understand how AI would interact with the network’s data and ensure that security and privacy could be managed at scale ahead of any potential rollout.

Work includes using currently available tools to enhance network efficiency, along with proof-of-concept projects to test more bespoke network use cases.

“One of the areas that we're experimenting in is using generative AI technology to mine data and insights from our notes fields,” Dutton said.

“In the future we'll be able to then apply that insight and information into our existing process automations and into the tools that we put in the hands of our technicians, but also critically provide that information back to the RSPs, which will then be able to go back to the customers and directly support that customer experience.”

The need to use AI would also increase over time, both as traffic increased, and as the network came under increased threat of disruption from climate-related events.

“The challenges that NBN Co faces in terms of driving that customer experience don't change,” Dutton said.

“Customers are going to want data when they want it, and they're going to want that reliable experience, and what we'll use AI and other emerging technologies for is to continue to improve that and get us up that maturity curve.

“AI and other emerging technologies on top of that will continue to allow us to drive that reliability.”