How should air navigation service providers (ANSPs), airports, and airlines introduce artificial intelligence (AI) technologies into their operations to exploit the performance improvements that AI offers while not disrupting legacy operations? According to industry leaders meeting on 23 May in Brussels at the inaugural conference on aviation and AI, which was organised by Eurocontrol and Airbus, one proven concept is to develop a ‘digital twin’ of a legacy operation. This means replicating an airport tower operation, airfield, or an entire aircraft in a parallel, digital model.
Andy Taylor, chief solutions officer at NATS, described how the UK ANSP developed a digital tower at Heathrow as a backup – and parallel – solution to the live operation. Apart from understanding how AI technologies can be seamlessly introduced into the operation, the digital tower also provides the Civil Aviation Authority (the UK regulator) with a clear understanding of the concept – whereas more theoretical models do not. This is a major advantage in the certification process.
The challenge of certification is one reason why the aviation sector has been so slow to adopt AI. According to Eurocontrol Director General Eamonn Brennan, only 4% of all AI patents are filed for aviation operations.
However, adoption of AI is set to accelerate. Airbus set up an AI team in 2015, and by 2017 it was managing 259 proof-of-concept programmes. According to Airbus Digital Transformation Officer Marc Fontaine, the projects include algorithms to convert broken-English cockpit conversations into reliable text messages, and aircraft sensor/object resolution systems to track the progress of turnaround operations on the ramp.
“Big data is the most important and fast-moving industrialisation technology we are looking at,” he added. “It’s important not to touch existing systems but find ways to extract data from them.”
Want to read more? For analysis on this article and access to all our insight content, please enquire about our subscription options at ihsmarkit.com/janes