How can AI technologies contribute to predictive maintenance in energy systems ? What impact does this have on operational costs and sustainability efforts ?
Sophie : Predictive maintenance is top of mind for me . Going back to Torresol as an example , I feel that this is interesting and something that not many people are talking about .
Continuous monitoring of the performance of energy assets is so critical in a world that ’ s diversifying its energy sources and can ’ t compromise on reliability , that ’ s critically important . To be comfortable with the provision and reliability of that energy source means tools like predictive maintenance are critical .
Kevin : Let ’ s say I want to monitor a transformer in the field . I can say to IFS that we have performed all the preventive maintenance , what ’ s the realistic lifespan and health of this asset at any given point in time ? What AI can do to help us there is make sure we ’ re providing the right maintenance , whether it ’ s predictive , proactive or even preventive on schedules that the manufacturer has dictated to us . And if I ’ m doing those things , am I optimising the efficiency of that asset but also the lifespan of that asset ? And that ’ s one of the key differentiators that IFS offers in this space .
The other thing we have when we look at the work that needs to be completed , if IFS ingests a sensor code that says that this is an X alarm fault , we not only know what that means , but we know which technician we should dispatch to it with the right certifications , we also can recommend which spare parts to send because it knows through the analysis of the data we needed these spare parts to fix it . This means technicians don ’ t need to be making multiple trips out to diagnose and repair . Every one of those pieces in that chain has the ability to either impact positively or negatively on consumption of energy within the sustainability chain .
For example , utility PG & E can , based on prior events , detect certain wind gusts and if they blow above a certain threshold , they turn the power off . This is to prevent fire from wind damage or other things that could potentially cause bad environmental impact and outages .
When they determine the wind has died down and they can turn the power back on , it takes hours for them to survey every line to make sure there ’ s no debris or interference . But if we are even able to dispatch drones for that inspection of the lines , that alone means we ’ re getting much more efficient in preventing what could be a catastrophic impact . Those are more of the real-world examples that I ’ m seeing from prospective customers and customers and how they ’ re managing or using technology to prevent issues .
66 February 2025