Uptake and Rolls-Royce are about half way through a two-month proof of concept to ascertain whether a different data-science approach will improve Trent engines’ reliability—including detecting problems before they occur.
Rolls-Royce has decades of engine monitoring and data analytics service experience from TotalCare, but reliability problems with the Trent 900s (A380) and Trent 1000s (787s) have cost it more than £1 billion (US$1.27 billion). By pairing its in-house expertise with Uptake’s industrial artificial intelligence and machine learning techniques, Rolls is hoping to increase engine uptime.
The proof of concept is not focused on solving particular operational problems. Instead, Uptake is using multiple data sets—including environmental and contextual information such as weather—to take a broader look at what could be contributing to failures—opposed to a more engineering and OEM focus.
By fusing different sets of information, Uptake thinks it can provide a fresh perspective on how operators are using the equipment or how susceptible engines are to particular environments, says Nick Farrant, Uptake's senior VP of business ventures.
“We’re not looking at ways to improve on-time performance on the day of flight, but rather, we’re looking at how we can enhance the models that [Rolls-Royce is] using to predict degradation,” he says.
To do that, as a starting point, Uptake is examining the Trent 700’s fleet data and is trying to characterize how the engines that power the A330 vary in use between regions. The next step is assessing environmental issues and challenges affecting the predicted life of the engine, says Farrant. The goal is to take this insight and use it to better predict the personalized care plan for an engine, he adds.
That could lead to increased borescope inspections, for instance, for a group of engines that fit certain operating characteristics.