The famous motto "Take the best that exists and make it better. When it does not exist, design it" of founder Sir Henry Royce is also a recurring theme in many areas of R² Data Labs, the global research centres of the global corporation, in the 21st century. In Dahlewitz in Brandenburg, near Berlin, Rolls-Royce has already been producing engines on a grand scale for 25 years with 2800 employees and is now expanding the site to include a new centre for Artificial Intelligence. The new hub is to act as an interface between AI, data engineering and Rolls-Royce's specific industry expertise, and will become a building block of Berlin-Brandenburg's innovative technology ecosystem. #KI_Berlin spoke to Dr. Klaus Paul of R² Data Labs about AI, digitisation and the future of aeronautical engineering.
Besides mobility, aeronautical engineering has traditionally been closely associated with the Rolls-Royce brand. Can you give us an insight into the AI innovations you want to initiate in R² Data Labs?
Right, it all started with cars. The aircraft engines were added a little later. Today we have had nothing to do with road mobility for a long time. What has remained is the claim of the first days, which runs from then to the current topic of digitisation and AI: "Pioneering the power that matters.” Wherever it matters, we offer proven and also new drive solutions.
The digital strategy of R² Data Labs has three pillars: we want to perfect our digital twins, create continuous data-based innovations, and evolve into a company in which digitisation is at the forefront. We started providing data-based services for aircraft engines two decades ago; now, with R² Data Labs, we are focusing on continuously improving our customers' operations in three areas. First, through greater efficiency: applications that help them manage their infrastructure and operations more efficiently and cost-effectively. Second, availability: applications that optimise fleet maintenance and management to minimize disruptions and maximize usable time. And third, risk and compliance - applications that support risk management and automated compliance processes.
Let's take the new "Pearl" engine generation developed and manufactured by Rolls-Royce in Brandenburg. What's so special about it, especially in view of real-time analysis of data and machine learning?
All Rolls-Royce engines can already be considered intelligent in that data play a role in their design, assembly and maintenance. Our new Pearl engine family has been the most successful of these to date - it is the latest example of the implementation of an IntelligentEngine.
In addition to a new Engine Health Monitoring System for function monitoring, the engine for the first time integrates a remote diagnosis function with bi-directional communication. This makes it possible to configure the monitoring functions of the engine from the ground: bi-directional communication allows the monitoring system to be adapted to the operating situation from the ground. In return, it sends all relevant data back to our control centre. In total, the electronic engine control system records up to several thousand parameters in parallel, about one third of which are permanently read and temporarily stored in the aircraft.
There are already some prototypes of swarm robots, small nano-robots which could further minimize errors and damage to engines. Are we on the brink of a revolution in engine maintenance?
When it comes to engine maintenance, I would rather speak of a continuous evolution, with the aim of combining the greatest possible safety with the same efficiency, availability and falling costs. A great deal of this is already being achieved today with our forward-looking maintenance range through CorporateCare® and TotalCare® contracts. In short, we obtain fleet operating data and use Big Data, machine learning and AI to compare and forecast the behaviour of one engine with that of all other engines of this type. The conclusions that can thus be drawn for service have become more and more significant - among other things due to ever larger data volumes per flight - from formerly kilobytes, the trend is now moving into the terabyte range.
The swarm robots mentioned are currently still too large to be released from the laboratory into reality - we will certainly be much further along in a few years' time.
Rolls-Royce operates worldwide and has expanded its location in Dahlewitz to include a centre for artificial intelligence. Why is Berlin-Brandenburg the right location for this and not, for example, the London headquarters?
Our R² Data Labs consist of a very small headquarters in London and satellites with expert teams in strategically selected locations - in the UK, Asia, the USA and now here in Germany. The satellites are closely integrated into our operational business units. They systematically strengthen their internal competence and capacity by means of external specialists. Rolls-Royce Germany is headquartered in Dahlewitz in Brandenburg. In Berlin, more or less in the geographical centre of Brandenburg, and in Potsdam, the state capital, there is a great deal of expertise that we can integrate into our ecosystem. Instead of a large central force, far away from the area of operation, we work with more small, agile units where the expertise is needed.
There are already many interesting AI research projects at universities, start-ups and other companies. What expectations do you have of Berlin as a business location, and have you already considered cooperation projects?
We are already cooperating with the Hasso Plattner Institute in Potsdam and have been working closely and successfully with the TU Cottbus-Senftenberg for years. Our proximity to Berlin will certainly contribute to broadening the spectrum. We are already talking about this with some companies and organizations - but not yet publicly. And we attend a large number of events, of which there are a lot in Berlin, in order to make new contacts and learn.
Take a look into the future: what impact will AI have on mobility and aviation in 10, 20 or 30 years?
The influence is sure to grow! We did not articulate the vision of the IntelligentEngine until 2018, for example - on the basis of both our experience and trend analyses of customer needs and the technical and methodological possibilities which are emerging. One thing is clear: our customers want to avoid service interruptions and reduce their costs. They want to make more and more environmentally friendly mobility offers. We can and wish to contribute to this, especially through AI. As an established manufacturer, we are in the privileged position of being able to access a large historical and constantly growing data pool from the development, construction and operation of engines. AI helps us to learn from this data and to identify relevant patterns for users, such as airlines. I suspect that we are not alone in our industry with this approach. So it remains exciting!
Thank you very much for the interview.