Changes in software algorithms are reducing the need for sensors.
Altair and Rolls-Royce Germany’s collaboration on aircraft designs now incorporates a higher degree of artificial intelligence (AI) and machine learning (ML). In Q3 2020, the companies signed a memorandum of understanding to integrate AI and ML into the engineering, testing, and designing of aerospace engines. The move will lower the number of sensors needed for tests, reducing development time and costs. A shorter design time will get products to market faster.
“We are using AI and (in designing) all of our products and services to unlock additional value. In this specific case, working with Altair, we’ll use the potential of AI across all of our business aviation programs. This includes our Advance2 technology demonstrator program, which is being used to develop and improve technology for future product generations,” said Peter Wehle, head of innovation and research and technology at Rolls-Royce Germany.
Wehle said that AI is also being integrated into processes to design models of Pearl engines, “which power some of the fastest, largest, longest range business jets from Bombardier (and) Gulfstream.”
The savings for measurement equipment could be in the millions of euros.
“There are about 5,000 parameters….within a typical engine development test campaign. [The] cost to measure [just] one parameter can vary between 500 and several thousands of euros,” said Wehle.
AI Informs Engineers About Performance During Use
The hidden value of AI is its ability to eliminate the numerous daily tasks needed to collect data and evaluate them manually.
AI also allows in-service data, data collected while an engine is in use, to be fed into the design process, said Sam Mahalingam, CTO of Altair.
“In 2020, our engineers at Altair put a lot of effort into fine-tuning the algorithm. This helps us identify and remove noisy data (data relating to meaningless information). These data do not contribute to improving the design of the craft. We can then take high-quality information, organize it, and shape it into a form,” said Mahalingam.
Rolls-Royce’s engineers can query, or request access to data about the form, using machine learning.
“Having access to information (like the amount of thrust the engine will generate) in early stages of design will lead to fewer iterations,” said Mahalingam.
Rolls-Royce Germany has relied on AI for over 20 years. Yet the company is now incorporating it and ML to a greater degree because of its vision for IntelligentEngine. This is a technology in which a jet engine is connected to and aware of other engines, its support ecosystem, and its customer airline. IntelligentEngine allows a jet engine to learn from its experiences to adjust its behavior and achieve best performance.
“Altair … came up with a disruptive meshless structural analysis tool (a tool that causes significant change in the market, analyzes the effects of loads on physical structures, and is based on the interaction between a communication endpoint of the engine simulation with neighboring points). [Altair also] added a range of data analytics tools to their portfolio. The next step shall be to connect those tools to unlock the full potential for extremely fast and deep insights into the structural system behaviour,” said Wehle.
Rolls-Royce Germany is Planning Ahead
Wehle said that Rolls-Royce Germany has a long-term vision of building a system-level design recommender system. A recommender system is an algorithm that provides the most useful information by filtering it from a larger database.
“We would like to give the non-data scientist engineers access to data science methods
within a software environment they can work in and that can be maintained. Ultimately this collaboration will help to democratize our data analytics, enabling our engineers to make better daily data-driven decisions. [That will] transform our business and products,” said Wehle.
Wehle noted that structured data will allow Rolls-Royce Germany to better share data between its different departments, including development, manufacturing, and service.
One of the ways that Rolls-Royce is currently sharing data is through R² Data Labs, a tech ecosystem through which the company communicates and works with start-ups, educational institutions, industry partners, public sector bodies, and customers. R² Data Labs includes a Digital Academy, which trains staff in digital skills. The Data Labs utilize The Aletheia Framework, a toolkit that scrutinizes the application of an AI to ensure that it is ethical.
“The Aletheia Framework also controls bias by a five-step checking process on the decision made by an AI. This allows us to trust its activities and demonstrate they are ethical,” said Wehle.
Wehle said that R² Data Labs focuses on three key areas, efficiency, availability, risk and compliance. Efficiency improves asset operation and maintenance. Availability maximizes the availability of assets to increase productivity. Risk and compliance strategies focus on improving risk management and automating compliance processes. Shifts in these areas allow Rolls-Royce Germany to improve operational performance for its customers.
Over time, the improvements that Altair and Rolls-Royce Germany develop are expected to filter down to the customers. The technological changes and new practices will help all parties work together more closely and effectively.
Altair’s Acquisitions Enhance Custom
Mahalingam said Altair has improved its ability to serve Rolls-Royce Germany and other customers in the aerospace sector. This is because it has acquired companies and strategic technologies that add value and engage in new and helpful tasks.
The most recent significant acquisition for its work in aerospace engineering is Datawatch, which became a part of Altair in 2018.
Datawatch was originally a U.S software company that developed self-service data preparation solutions. Self-service data preparation involves the creation of tools that enable users to access, merge and manipulate data sources without the assistance of IT staff.
Mahalingam said that Altair is also assisting other aerospace manufacturers with its software.
“All of the changes we are making in software is use-case specific. Most Machine Learning models are created for one type of engine,” said Mahalingam.
Since aerospace design is extremely competitive, Altair signs a contract with each customer to ensure that the ML models become the intellectual property of the customer.
The Pandemic Hasn’t Reduced Demand
The pandemic has not negatively affected Altair and Rolls-Royce Germany’s interactions.
“We were already prepared to interact remotely because our tools are being offered on the cloud. At certain points, if the pandemic had not occurred, we would have met face to face. Yet not being able to do so in the past year has not impacted the partnership,” noted Mahalingam.
Mahalingam said that the pandemic-related drop in air travel has not reduced the demand for Altair’s software.
“Companies have used the time to expedite and accelerate the design process. They’ve been able to hold many conversations with us about how to bring the data in different silos (collections of data) together,” said Mahalingam.
Mahalingam said working in the cloud has made it easier for Altair to assist Rolls-Royce Germany during the design process.
“If they’re still modeling in 3D, and we’re creating the specific ML model, one of their engineers can say, ‘Can I submit this design?’ The AI and ML make it possible to introduce an additional feature or innovation. Then we can plug it into our tools without disrupting their workflow,” said Mahalingam.
