A prediction on the role of AI for an aviation industry in crisis

David Perez

2020-05-06 14:02:59
Reading Time: 3 minutes

The coronavirus pandemic is making many industries rethink how they introduce their services to the public. Aviation, in particular, will be impacted for a variety of reasons. As a means of transport, flying is the most incompatible with social distancing, which would be difficult, expensive and almost impossible to implement. While some airlines have announced that they will leave middle seats unoccupied, implementing this change will undoubtedly raise ticket prices.

Additionally, corporate travel will immediately change after the pandemic; the impact that will have on aviation as a whole is difficult to foresee in its entirety. There is also uncertainly surrounding the internet and its relation to the travel industry due to its increased usage. While less spending and vacationing will surely impact the industry, a beach experience cannot exactly be replaced; in contrast, meetings can easily be simulated with modern videoconference equipment, electronic document management systems, etc. In any case, as fewer travel, travel costs will decline as airlines eagerly reduce prices as a way of recouping losses – these prices are expected to increase over time in following years. Though many are already discussing a path to rapid recovery from this pandemic, the previous crisis impacted the industry for at least half a decade. Our team is developing a set of descriptive analytics tools to efficiently monitor air traffic and extract conclusions about the evolution of the air transport network. You can read about it here.

Artificial Intelligence technologies were pivotal in roadmaps to help aviation across many fronts. Challenges that seemed too big to be covered by any other alternate technology, like runway capacity management now have seemed easily accommodated in recent years. Assisting and predicting things like runway occupancy times, equipment approach and tower controllers was one of the first applications of AI. Other applications in aviation, like platform management, will be more easily implementable with decreased operations, which will likely make potential delays and disruptions less of an issue.

In the past, efficient fuel uplift and usage was a very promising application of AI tools; now, however, it seems like this will shift to being low priority in a scenario of incredibly low oil prices. Efficient routing will also shift in priority, as slot constraints will not be an issue for airlines as they ramp up operations – industry sources are expecting only 40-50% of their operations to recover by the end of 2020.

The passenger experience will probably shift to safety and extensive use of biometrics; streaming the security process through new AI technologies and other methodologies would require investments by airports that will not be available.

New aircraft capabilities automating certain pilot functions, including automation of the cockpit, will be postponed, especially considering the precarious financial situation of Boeing and Airbus. This, together with the fiasco of the introduction of heavy-IT systems in the 737Max will, very likely, delay the introduction of AI in the cockpit.

Air Navigation Service Providers (ANSPs) will also face at least a few years of cost-control and petitions from the airline industry to reduce ATC fees. This, along with lower traffic, may pave the way to new AI systems that take automation of air traffic control services to the next level. Assisting controllers to manage low risk sectors or even resectoring with automation in mind will be doable now – the same scenario and development might not have been advisable (or necessarily from the cost perspective) with higher levels of traffic and complexity. AI could assist in defining better sector configuration and supporting ATCOs, if not outright automatically controlling air traffic thanks to CPDLC. ANSPs leading AI technology will become more cost-effective and will gain the necessary experience in automation to address more complex challenges and beat capacity limitations when pre-pandemic levels of traffic come back.

While the aviation industry settles and starts rebuilding their economy, we can also rely on data science to provide analytics and predictions on traffic evolution. Have a further look!

Author: David Perez

© datascience.aero