Origin Energy is turning to artificial intelligence (AI) and automation to help its network team shift away from performing manual and repetitive tasks and add greater value to its operations.
The company’s head of networks and core infrastructure services Oliver Oldham said his team has been converting processes that previously involved paper forms, Excel spreadsheets, command line interfaces, and manual configuration changes and automating them as software.
The result has been that his team no longer spends as much time capturing and analysing packets to resolve ‘housekeeping’ activity, which Oldham said meant they were free to focus on more valuable activities, projects, and strategic work.
“We’ve started to develop AI, which has allowed us to integrate that into a lot of the low-level tasks around service optimisation,” Oldham said.
“This has allowed us to eliminate a lot of the errors that would creep into running network services and eliminate the low-level incidents and problems that would work their way up to the network operations team.
“This allows us to focus on driving more efficiency, servicing our business stakeholders, and delivering more efficient network services, which results in less calls to the service desk, and at the end of the day, happier end users.”
The program has examined numerous end-to-end network change processes, which are coded into software and then delivered through a continuous integration and continuous delivery (CI/CD) pipeline.
“It allows us to automate and test and to deploy that code into an environment that we know the configuration state of, and it removes a lot the intricacies you have when you have multiple individuals taking part in that change process,” Oldham said.
“We’ve gone from an environment where we would see around a third of our network changes run into problems, to a point now where less than five percent of network changes are unsuccessful.”
Oldham said this program had also led to a more stable network environment.
“So all of those metrics together really paint a picture of a more efficient, stable service, and happier users,” he said.
While for now Oldham and his team have been content with allowing the software-based system to manage low-level activities, he said he was also keeping a watch on how these technologies evolved.
“The words AI and self-healing are something that really has created quite an incredible marketing buzz at the moment,” Oldham said.
“If I look wider than just that network space, we’re seeing stuff coming out from some of the larger technology players around Copilot and generative AI companions.
“That really boils down to just taking care of that low-level admin and time-consuming tasks, and giving individuals quicker access to the information they need to make the bigger decisions.”
To date however he is happy to continue automating what he describes as the ‘low hanging fruit’.
In a recent example Oldham’s team was able to quickly identify a fault within its KVM (keyboard, video, mouse) hubs that was causing intermittent switching between wired and wireless networks, interrupting video calls.
“That was something we were able to resolve almost automatically through the system, which traditionally would have taken someone maybe a few days to contact the users, try and replicate the fault, capture the data, and understand what’s happening,” Oldham said.