The aviation industry is under pressure to improve the sustainability of air travel while improving operational efficiency in an increasingly complex marketplace that is still recovering from the impact of the COVID-19 pandemic. In an industry where safety is paramount and new technologies require utmost scrutiny, generative AI promises to boost aviation businesses and their industry partners.

There are a myriad of potential use cases for generative AI. Some use cases require time to integrate with existing business systems and processes, but industry leaders should move forward to areas best suited for generative AI’s strengths. These include enhancing customer experiences through better, more personalized travel recommendations and promotions, as well as improving customer service by creating more robust virtual assistants. Generative AI could also help maintenance, repair and overhaul (MRO) technicians by enabling them to retrieve relevant information more effectively for repairs, or by automating the creation of parts and equipment orders so repair or maintenance can start as soon as a plane lands.

Some airlines are already using generative AI in communications and customer service operations, including automating translations of texts, producing marketing materials and writing copies. AAR Corp, a private provider of aviation services, is considering the use of generative AI to optimize inventory management, provide predictive maintenance, improve warehouse operations and automate parts ordering.

The aviation industry is prudently taking a cautious approach to rolling out AI. In May, the European Union Aviation Safety Agency (EASA) released the second version of its AI Roadmap (link resides outside ibm.com), which provides a comprehensive plan for the integration of AI in aviation, with a focus on safety, security, AI assurance, human factors and ethical considerations.

Such measures indicate that the industry is aware of AI’s risks, which include bias in the data used by machine learning systems to train AI models that can skew recommendations or analyses. In addition, generative AI can sometimes identify patterns or objects in data that are nonexistent and yield results that are either nonsensical or altogether inaccurate—a phenomenon known as a “hallucination.”

Because of the possibility of error in generative AI, aviation companies should begin by exploring use cases where some variability can be tolerated, such as customer service bots. They must also make sure that customer data is secure and that its use is compliant with data privacy regulations.

With a focus on safety, data integrity and security, the industry can move forward with testing more advanced generative AI implementations. As aviation companies integrate generative AI with their existing intellectual property and systems, they will unlock the technology’s value in a number of ways (link resides outside ibm.com).

• Personalize pilot training: Generative AI can take current training simulators to a new level of realism while modifying training scenarios to each pilot’s performance.
• Optimize flight routes and schedules: The technology can also merge data from historical and real-time flights, weather patterns and air traffic to find the most fuel- and time-efficient routes and schedules, minimizing delays and fuel consumption.
• Support predictive maintenance: Generative AI can analyze data from sensors and other sources, comparing it to historical data to predict potential failures and optimize maintenance schedules.
• Create digital twins: Aircraft designers and MRO companies frequently create virtual models of products or systems. These digital twins let them test designs and drive performance more quickly, without the need to produce physical models. Generative AI can build digital twins much faster than current methods.
• Plan and optimize airports: Complex digital twins of existing or new airports allow designers and planners to simulate and boost passenger flow and aircraft ground traffic, as well as building infrastructure and amenities.

These examples merely scratch the surface of possibility, as different players in the aviation industry might enlist generative AI to help with aircraft design and prototyping, supply chain optimization, advanced flight planning and more.

Ultimately, aviation companies must find the middle ground between innovation and caution on the path to becoming industry leaders in AI. With so many use cases, generative AI is ready to transform every aspect of the industry.