Airbus accelerates innovation with Azure Cognitive Services

Airbus looks back proudly on 50 years of innovation in the European aerospace industry, but its focus is on the future. The company is a leading producer of aircraft and helicopters for both commercial and military customers, plus launch vehicles and satellites. It also delivers navigation, secure communication, data services, and other solutions to its global customers. As the company regarded two very different use cases, it found that Microsoft Azure technologies, especially Azure Cognitive Services, offered the best path forward. The aerospace pioneer continues to push technical boundaries to help make a better-connected, safer, and more prosperous world.

Meeting varying and strict security needs

Many Airbus customers, especially those in the military and government sectors, must comply with highly restrictive regulations that preclude the use of public clouds. A number of countries have strict data nationalization rules that also dictate a non-public option. “We’re also governed by a different set of regulations for our work with organizations like NATO,” explains Marcel Rummens, Product Owner of Internal AI Platform at Airbus. “When you consider the vast number of regulations we must comply with, especially in Europe where many countries have their individual rules, you can easily see the difficulty in bringing modern services into this environment.” 

The company turned to Microsoft for the technology it needed to build its restricted cloud while maintaining compliance with data sovereignty regulations. Part of its platform offering are Azure AI solutions. “Only Microsoft offered the proven, cutting-edge technologies and models that have been trained on Microsoft data sets on a very large scale,” explains Rummens. “Microsoft also makes these technologies available in disconnected (removed from public cloud) mode. All of this accelerates our time to market, and that has been the key differentiator for us.”  

Training pilots easier and faster

Pilots training to fly new aircraft have always faced a rigorous regimen of classes and testing. But as aircraft gain more capability, they become more complex, escalating the volume of pilot training material to more than 6,000 pages of technical information. Pilots must master it all during intensive multiple-week courses and recall it accurately for as long as they are rated (certified) for that aircraft. 

To fill the gap between the initial classroom training and twice-yearly refresher courses, Airbus created a pilot training chatbot that supports trainee pilots to self-guide through the material. Pilots access the chatbot, deployed as a mobile website, on their mobile devices for fast answers to their spoken questions. Even at this early stage, the pilot training chatbot garners appreciation among the test group. “The biggest challenge for our pilots is maintaining their level of familiarization with complex aircraft systems for the duration of their flying career. The training chatbot is almost like having an on-call instructor available anytime our pilots need,” says Peter Walker, Project Lead for the Airbus A330 Pilot Training Chatbot. “It resonates with digital natives, of course, but regardless of age, no one wants to carry thousands of sheets of paper around.”

Because Airbus also created pilot training chatbot versions for military aircraft, in contrast to the civilian aircraft version, the company needed to isolate it from the public cloud. To achieve the disconnected, “air gapped” deployment required, Airbus deployed speech containers via Cognitive Services on its own Kubernetes cluster, which is part of its restricted cloud. Stephen Gerard King, Data Analytics and Applications Analyst at Airbus, joined the Speech to Text and the Text to Speech containers in Cognitive Services with the Speech SDK, which they used to create the speech-enabled application. “The Speech SDK was very clear and straightforward, which greatly accelerated the integrations,” he says. “The Microsoft product team was also very responsive.”

Supporting military aircraft 

Managing the complexity of military aircraft flight operations demands data-intensive precision—a perfect AI technology application. During flights, aircraft are put under many different types of stress. This is especially true for military machines as they have to operate in even the harshest of conditions. Factors like rapid changes of temperatures, radiation in high altitudes, or just normal wear and tear can lead to problems impacting the operational readiness of the aircraft. 

To better understand the health and condition of the aircraft and foresee and fix potential problems before they occur, Airbus deployed Anomaly Detector, part of Cognitive Services, to gather and analyze the telemetry data. It began a proof of concept of the aircraft-monitoring application by loading telemetry data from multiple flights for analysis and model training. “For more complex cases, we’ll have to develop custom algorithms,” says Rummens. “But early tests have shown that for many cases, the out-of-the-box solution works beautifully, which helps us deploy our solutions faster. I would say that we save up to three months on development for our smaller use cases with Anomaly Detector.” 

Flexing to meet the need for disconnected solutions

Rummens’s team deployed Anomaly Detector in containers. They are also experimenting with both the single variant and multivariant Anomaly Detector, and they look forward to their next phase: using translation APIs in Cognitive Services to overcome the language barriers between countries using Airbus systems, especially for high security applications that preclude the use of a web-based translator like Bing. 

The team needs to focus on the forward-looking solutions they’re innovating, not on infrastructure. “The Microsoft solutions just work,” says Rummens. “We also prize the disconnected solutions. On top of that, Microsoft adds new services frequently and constantly updates and enhances existing services." King agrees. “We have tested different cloud APIs that are very easy to use,” he says. “These are a disconnected equivalent, making the Cognitive Services solutions both simple and fast to deploy for our proof of concept.” 

For Airbus Head of Digital Platforms Alexander Kraehenbuehl, the new applications open possibilities for the company. “There are recurring analytics challenges that you do not want your scarce data science resources to solve over and over again. I am pleased that thanks to our close collaboration with Microsoft, we will be able to offer reusable AI services on our restricted cloud. And this can be fully disconnected when required by our government customers.”

Find out more about Airbus on Twitter, Facebook, and LinkedIn.