Washington product intelligence

Know that moment when you can’t find your keys, or your glasses, or your phone charger? That frustrated can’t-get-on-with-life feeling?

Now imagine it on an industrial scale, right across a factory floor. At the Rolls-Royce site in Washington, UK, they have a Product Intelligence system that takes those frustrations away, creating new levels of efficiency.

Under pressure

The 18,000 sq ft plant manufactures discs for a variety of Trent engines, taking a moulded alloy and machining slots along the edges with a “fir tree” design so that turbine blades can then be slotted in. Each disc needs to be machined and treated precisely, to withstand the heat and pressures inside an engine, and also to ensure the blades stay exactly in place, given that the forces on that fir tree slot can be about 100 tons or the equivalent of hanging a freight train off each blade.

Production intelligence

With each disc costing up to £100,000 and a challenging production schedule that demands 2,500 of them are produced each year, it’s vital to know exactly where each and every one of those discs is, and where it stands in the process, to maximise the production flow.

Despite being up to 160kg in weight and as wide as a car tyre, it is still surprisingly easy for a part to go “missing” if it is sent back through the production process for additional work and has to take its place in the queue along with the rest of the workstream.

That’s where Production Intelligence comes in. Each disc now has its own router with its own part number, sending out a signal 24/7 that connects it with a GPS map of the plant. If any production leader needs to know exactly where a part is, they can log into a specialist app on their phone – and find out where it is.

At the site’s central hub, production leaders can take that information a lot further. All those transponder signals are shown on one large screen, displaying a wealth of data on each disc’s status – how long it has remained stationary in the system, how it stands against its production dispatch target (it typically takes 4-6 weeks for a disc to make it through the entire system), and engineers can tag on their own comments about any issues that may be hindering progress.

Which means team leaders at the hub can then see the overall picture and prioritise the order in which discs are machined.

Keeping to the plan

Ryan Malley, Manufacturing Manager at the site, said: “Production Intelligence is hugely important to us here in Washington. We are a highly mechanised, automated plant but without this system we could still have team leaders spending hours of their working day simply walking around the plant tracking down an individual part and trying to find the background to any issues around it. Jobs don’t always flow in an orderly manner and its possible to lose scope of them and spend hours finding them in the facility."

“With this technology, we can set ourselves targets and check that discs gone through particular gates against a plan, allowing us to prioritise which discs need attention to move them through. Because we have all the information to hand we can have really dynamic meetings and we can agree our actions there and then."

“We can even trace discs hundreds of miles away in Derby if they are moved there for an additional process such as welding."

“We’ve been fully operational at the Washington plant since 2016 and when we built it we invested around ten per cent of the total on IT, which is relatively high, but we knew that digital technology was going to be hugely important if we were going to succeed.”

Growing intelligence

Growing intelligence

The system is part of the Rolls-Royce IntelligentEngine vision where its products and services are increasingly connected, contextually aware and comprehending through digital technology. That vision also applies to its own operations.

And there is another, often unseen, element that Washington is working on – paperwork. Each part needs to be “signed off” before it leaves with confirmation that it meets stringent quality standards. The team is now looking at how that paperwork can be drawn into the system, so that they can also check the documentation is approved on time.

The Washington plant has more plans to use artificial intelligence and machine learning to further improve efficiency. Large electronic boards hanging from the ceiling of the plant currently show the status of each machine and when the part it is working on is expected to be completed – but the team are looking at how the machine could also signal to the board what tooling will be required for the next job, and if there is any help it needs.

“We are moving to an AI system where our hub will progressively learn from the operations we book into the Production Intelligence system and, through its understanding of the workload and targets, create a plan for each of our machine operators that constantly is updated and adjusted,” said Ryan.

Load more