IIoT and AI Provide the Fuel for Industrial Efficiency at Rolls-Royce

Rolls-Royce's industrial IoT initiatives have created a big, new challenge for itself: managing the data derived from 70 trillion data points. To do that, it turned to artificial intelligence and is now sharing a framework and toolkit to help other manufacturers apply—and trust—AI.

Image credit: Rolls Royce

Jet engines aren't cheap, not with a price tag that runs anywhere from $12 million to $35 million. And with that threshold of customer investment, Rolls-Royce plc decided to invest in new ways to monitor and troubleshoot its turbine engines, significantly bolstering the safety and reliability of its products, according to Nigel Hart, Lead of Digital Transformation at Rolls-Royce's R2 Data Labs.

"The value promise of that asset allowed us to use what was at the time specialist and high-end technology, long before anyone was talking about the Internet of Things," he says. And it was from IoT that Rolls-Royce customers have been able to perform advanced monitoring and troubleshooting, saving them downtime hours and unforeseen spending in an industry where safety and reliability are always the top priorities.

Hart explains that the manufacturer has been testing and using what's now widely called IoT technologies for the past 20 years. More recently Rolls-Royce has also watched as IoT components and connectivity have decreased in price, making the technology even more attractive.

"The cost of acquiring [IoT] data has also decreased, which allowed us an organization to do two things: Collect data from a broader range of our products, and connect more of our assets--machines, equipment and products," Hart says. "As costs come down, the scope opens up."

Two people walking in Rolls Royce manufacturing floor

Image credit: Rolls Royce

Managing Trillions of IoT Data Points

While that's all good news for the company's focus on safety and innovation, Rolls-Royce's industrial IoT initiatives have created a big, new challenge for itself: Managing the data derived from 70 trillion data points. In addition, the number of data points will continue to grow over time, not to mention the volume of data they'll generate–-an embarrassment of riches.

That, in turn, has forced the manufacturer to consider its IoT rollout a little more carefully. "Just because you can connect something, does that mean you should?" Hart asked. In these instances, smart, successful companies come back to the measure of what the value of IoT is to the business. "We're now looking to replicate that model from aerospace across our business," he added.

Rolling Out More IoT Solutions

Consequently, Rolls-Royce did that math and extended its industrial IoT rollout to its MTU business unit, a diesel power manufacturer based in Friedrichshafen, Germany, with offices in 175 countries.

"By developing a much lower cost IoT technology than we used on our aero engines, we can now connect MTU customers' diesel engines," Hart says. "It's a vote in the technology but also for greater business awareness. IoT allows us to optimize and build value services, not to mention run the business more effectively," he adds.

With these industrial IoT rollouts have also come some hard lessons for the engine maker. Top of mind for Hart is that IoT implementers must consider the total costs of implementation. "It's not just the sensor and the device, it's the connectivity and the solution," he says. "And those bills need to be paid over time. You don't just jump from the box to activity" without paying as you go.

Energy Management of IoT Devices

Rolls-Royce has also spent a lot of time trying to understand effective energy management for its IoT devices, particularly with batteries. "If there's an asset, I want to connect it in one job and walk away. I don't want to have to come back to it," Hart says. This is not an insignificant objective when more and more IoT devices are getting smaller and smaller, even microscopic.

Another challenge for industrial IoT implementations like Rolls-Royce's is managing what Hart calls "the tsunami of data" that results when millions of machine parts are metered. As that volume increases, IoT users must look at automated analysis and interpretation. As Hart wryly notes, "There aren't enough engineers in the world to monitor and manage what's going on."

Which is how Rolls-Royce and other IoT implementers opened the door to artificial intelligence and more automated review of data files. "Especially in safety-critical applications, how do we gain confidence that those automated processes are trustworthy and reliable?" Hart asked. "Going forward, we're going to be more and more reliant on AI-generated interpretation just to deal with and manage the much increased level of connected devices in our businesses."

Building Trust in AI

In that vein, Rolls-Royce in December 2020 published its Aletheia Framework™ which starts to define how an organization can ensure the trustworthiness of its AI equipment and data. The document is free to download and Rolls-Royce hopes it will be taken up by others looking to use AI as part of their IoT deployments.

The Aletheia Framework includes a 5-step process for "proceduralizing" AI ethics. Hart describes it as a checklist of questions that solution integrators, risk owners, board members, and other stakeholders can answer about the impact of AI in this context, not only inside the company but in the world at large as well.

Does that mean IoT and AI are inseparable? Given the volumes of data generated, Hart says yes. “I think that's accurate because of the volumes. To retain the trust of our customers and our customers' customers, we do need AI to progress," he notes. "Rolls-Royce operates in safety-critical applications, and we can't afford to fail."