Rolls-Royce managed close to 1,500 aircraft engine shop visits in 2025. As part of its successful multi-year transformation, of which digital and AI is a key part, its Optimiser AI multi-agent is helping teams navigate millions of variables faster, returning engines to airlines sooner, and keeping aircraft flying for longer.
When you settle into an aircraft seat, you probably don’t think much about the engines beneath the wings. You simply expect the aircraft to take off, arrive safely and, ideally, on time.
But behind every flight sits an extraordinarily complicated maintenance operation.
Rolls-Royce designs, manufactures and supports engines for civil aircraft. These engines are highly complex machines, made up of around 30,000 parts, with scores of sensors monitoring 100 parameters every 1.6 seconds, explains Matt Ley, Director of Engineering, Technology & Safety in Rolls-Royce’s Civil Aerospace division. This creates millions of variables across an annual maintenance programme.
And the difficulty is not simply repairing an engine; it’s getting it to the right place with the right parts, says Jamie Barkley, who leads aftermarket optimisation at Rolls-Royce.
“We have to weigh up capacity, capability, regulatory approvals and geographical challenges with logistics,” he says.
It’s a constantly shifting puzzle that consumes thousands of hours of planning time a year, says Barkley, much of it laborious and manual.
This is the huge challenge Rolls-Royce is tackling with an agentic AI system it calls Optimiser, built with Microsoft.
Agents working together
The Optimiser multi-agent is designed to make that planning quicker, easier, and much more responsive.
An Engine Agent draws on information about individual engines. A Workscope Agent helps determine what maintenance may be needed. Facility Agents consider workshop capacity and capabilities, while a Supplier Agent deals with the supply chain. The Optimiser brings those perspectives together, while people retain oversight and accountability.
Rolls-Royce has already moved more than 40 business systems into Microsoft Azure, providing a centralised data store that AI can analyse, adapting the maintenance planning as circumstances change.
“These agents allow us to get better and move closer to having an iterative process,” says Barkley. “When something changes, we can continuously optimise that plan.”
A planning exercise that previously required four or five people working together for about a week has been materially reduced.
Once the engine is in the workshop, the real work begins: taking the engine apart, assessing the condition of its components, deciding what can stay, what can be repaired, what must be replaced, then rebuilding and testing it, Ley explains.
After reassembly, the engine runs on a test bed and must meet specified safety criteria assessed by the regulator before it is cleared to return to service.
‘Happier customers’
For an airline, an engine awaiting or undergoing maintenance is an engine that is not powering a revenue-generating aircraft. At an industry level, shortages of available engines can leave aircraft sitting on the ground.
“The industry has sometimes struggled with ‘aircraft-on-ground’ – literally not enough engines available for the aircraft to keep all the aircraft flying,” Ley explains.
That problem became particularly acute following the Covid-19 pandemic. Passenger flights resumed quickly, but the aerospace supply chain, including its skilled workforce and parts availability, took longer to recover. Geopolitical instability also impacts supply chains for the industry, he says.
So, faster planning can help Rolls-Royce have capacity and parts ready in time for fitting, make maintenance schedules more predictable, and enable the company to return engines to airlines more quickly. This means more available aircraft and predictable fleet planning for customers.
Combine improvements across the processes “and then suddenly you’ve got much happier customers, you’re keeping all the aircraft flying,” Ley concludes.
The Optimiser is still developing, but initial results seem promising, with Rolls-Royce saying it is helping to reduce engine maintenance turnaround times.
With global demand for air travel continuing to rise, AI-powered efficiency gains like this can only be good news for airlines and passengers alike.