SA Power Networks (SAPN) has commenced a project with artificial intelligence (AI) software specialist, Complexica, to improve the accuracy of forecasting potential equipment failure across its asset base.

The project is aimed to assist SA Power Networks in optimising its operational expenditure on equipment maintenance, as well as capital expenditure on asset replacements, to minimise potential failure before replacement or upgrades.

SAPN, noting its extensive network of electrical assets across the state, stressed that technologies like AI would help to understand the optimal time to replace these assets and allow the firm to maintain its network in the most cost-effective way.

Complexica’s proprietary AI/ML engine focuses on forecasting the probability of failure of equipment, it said, and helps optimise replacement decisions by focusing on work orders with a high probability of failure.

Instead of manual analysis of historical replacement data, audit data and other datasets as well as external factors, the AI/ML engine would automate the analysis using a variety of smart algorithms which collect data and perform certain tasks such as augmentation, predictive modelling, trade-off analysis and optimisation.

“We are particularly excited to work with SA Power Networks within this scientifically challenging area of predicting asset failures and optimising maintenance and replacement decisions – the problem is not an easy one to solve, and [it’s] a perfect application area for advanced AI/ML methods,” said Zbigniew Michalewicz, chief scientist of Complexica.

“I look forward to personally working with the SA Power Networks team – along with my fellow AI/ML scientists at Complexica – to help enable their vision of optimised decision-making through advanced capabilities for prediction, optimisation, and simulation.”

SA Power Networks is the sole electricity distributor in the state of South Australia, delivering electricity from high-voltage transmission network connection points to around 900,000 homes and businesses.

The company is engaged in the management of South Australia’s electricity distribution network, operating over an area of about 180,000 square kilometres, and the provision of construction and maintenance services to the electrical infrastructure industry.