NBN Co is hoping that advances in artificial intelligence and machine learning can reduce the time to resolve incidents, and make the company more proactive in recognising outage conditions in the first place.
Executive general manager of technology transformation Jo Dutton told the CommsDay Wholesale Congress last week [pdf] that NBN Co is creating foundations for future AI work through its ongoing enterprise simplicity program.
The program has a data consolidation pillar of work, but little is known outside of that.
NBN Co has had AI/ML ambitions for some time, revealing work in 2018 aligned to IBM to use machine learning to aid pattern-matching and correlation of incident data.
Some years have passed, and AI has advanced in its capability over that time.
However, Dutton indicated that simpler systems on NBN Co’s side would allow it to unlock gains from AI/ML technology.
She cited an example of a customer that experiences a “mysterious issue with their internet connection” today.
“They reach out to their retail service provider and the RSP contacts us. We send out a truck to investigate and resolve the problem. The estimated resolution time is around 71.5 hours or approximately three days,” Dutton said.
Dutton is hopeful that the company set up AI/ML “to take our unstructured network and operational data to find patterns, predict and detect events” and to proactively notify customers there’s an issue with the NBN network “before they contact their RSP.”
She also said the ability to drive even further down the proactive detection and remediation path is “closer than you think”, though did not put a timeline on it.
She indicated a desire to utilise AI “to see degradation before the issue occurs [and] proactively fix the degradation using virtual assistants”.
“Supported by robotics, the problem is resolved quickly and efficiently before the customer even notices,” Dutton said. “That is our destination.”
Dutton also predicted that generative AI could support “everyday activities” at NBN Co, “such as helping technicians in the field and strengthening network resilience plans.”
