SITA's chief data science and AI officer Jean-Paul Isson has penned the first of what will be a series of blog post introducing us to concept and explaining why AI is "super-human at narrow tasks". Future posts over the coming months will look at AI applications within the passenger and bag journeys, airport and airline operations, ground handling and border management.
We hear a lot about the value of AI - NASA has even used it to discover two new exoplanets - but do we truly understand how it works? Mr Isson describes it simply as "a field of computer science that makes machines smart."
Effectively it is a machine powered by an algorithm used to solve a specific problem or complete a task that used to be handled by humans". "These powerful algorithms can process huge amounts of data, and they learn automatically from features and patterns in that data. They apply that knowledge to new inputs to perform human-like tasks," explains Mr Isson.
It is strongest working as 'narrow' or 'weak' AI which focuses on carrying out a specific task. 'general' or 'stronger' AI, when we achieve it, would also exhibit consciousness, along with the ability to think for itself based on self-awareness and genuine intelligence to match human intelligence. "Today's examples are still narrow AI, because they're programmed to complete tasks," says Mr Isson.
The digital revolution is helping drive AI, but the technology has been around in the aviation sector for more than half a century . "Automated systems have been part of aviation for quite some time - auto pilot and flight management, for instance. And now we're seeing drones and autonomous aircraft," says Mr Isson.
He describes the industry's digital transformation as "of vital importance" to AI's rise in air transport. "This is the most valuable process today for our industry, as has been the case in other industries," he says, that will "transform the business model of airports and airlines". Its key components - people, process, big data and more importantly technology - will sharply impact the industry, he explains, and adds that leading industry players are leveraging AI to harness the big data that derives from digital transformation.
"This is becoming front and centre," he says. "We in aviation must fasten our seatbelts! As AI inevitably takes off in our industry, it brings the potential to revolutionise how air transport of the future will operate. It's already disrupting the way companies approach their data, revenue management and entire operations," Mr Isson explains.
Forward-looking organisations in the industry are already using AI to reduce cost, improve flights and airport operations, enhance the customer experience, and achieve operational excellence. "The entire industry is poised to undergo unprecedented disruption," according to Mr Isson.
At a recent SITA Innovation Forum where more than 100 airlines and airports focused on innovation in air transport, the company says that almost every single attendee planned to invest or leverage AI to address their key business challenges, from airport operations, airline operations, bag management, ground handling and the overall passenger experience.
This reinforces the findings of SITA's latest Air Transport IT Insights survey, in which the IT investment plans of airlines suggest that AI will become commonplace. Over half of airlines said they expect to have AI implementations by 2021. In that same timeframe, the survey found that nearly 80% of airports will be using AI for predictive analysis to improve operational efficiency.
Mr Isson outlines further research that supports this. According to a survey by NewVantage, 97.2% of companies said in 2018 that they were investing in big data and AI. A forecast from Research & Market put the global AI market in aviation at USD2.2 billion by 2025, posting a 46.65% growth (CAGR) compared to 2018.
SITA itself is making a significant investment in this field with the creation of the SITA Centre of Excellence in Data Science and AI in Montreal, an AI Hub where the company is planning to hire around 60 data scientists. Using AI and data science, it aims to help the company, its customers, partners and the entire industry to create actionable business value from the air transport industry's data assets.