KLM Royal Dutch Airlines will collaborate with Utrecht University in the AI Labs in the field of artificial intelligence. In the AI & Mobility Lab , young scientists will spend the coming years researching how smart software can support KLM in robust operational planning and recovery from disruptions.
KLM operated 169,000 flights last year, transporting around 30 million passengers. A significant number of passengers have a transfer to another flight at Schiphol Airport. These numbers are expected to continue to grow in the coming years. Successful and timely completion of each flight requires multiple resources to come together at the right time and in right numbers. The planning of this resources is becoming increasingly complex, especially in the face of uncertainties around weather, employee sickness and equipment failure.
Associate Professor Marjan van den Akker of the AI & Mobility Lab : "To operate a flight, a lot is needed: of course an aircraft, cabin staff and a pilot, but also all kinds of activities on the ground such as loading and unloading baggage, desk agents and catering. This means that if a flight is delayed or an aircraft breaks down, other processes are also disrupted, such as baggage handling or passengers and crew missing their connection. Artificial intelligence can help to effectively and quickly alleviate such disruptions."
The most data-driven airline
"At KLM, we are committed to innovation and this partnership underscores this commitment", says Bob Tulleken, KLM’s vicepresident Operational Decision Support. "We strive to become the most data-driven airline and this collaboration with Utrecht University exemplifies our dedication to staying ahead in a challenging industry. By leveraging AI, we can build more robust and adaptable airline operations."
Due to its expertise in algorithm optimization and experience in the field of aviation, transportation and logistics, Utrecht University is ideally suited to tackle these issues together with KLM, Van den Akker agrees. "The knowledge of our researchers and the wishes of KLM are very closely aligned. Within the AI Labs, we can also exchange knowledge and experience in the field of artificial intelligence with the other partners. NS in particular faces similar challenges as KLM does."
The knowledge of our researchers and the wishes of KLM are very closely aligned
Marjan van den Akker, Associate Professor
Two PhD students are researching under Van den Akker’s supervision on how KLM can use smart technologies to adjust planning at the last minute. "At KLM, they now work with mathematically oriented models for separate fleet and crew management. This means that each department has its own planners who only look at the fleet, or only at the crew and how to shift them. However, as fleet and crew resources become more constraint and disruptions more common, better decisions can be made by jointly optimizing for fleet and crew."
To do so, one of the PhD students, Philip de Bruin, is now developing an algorithm that can quickly calculate whether you can shift pilots and aircraft, or - if that does not work - see which flights are best to delay or cancel while minimizing impact on passengers and crew. This integrated approach allows more efficient recovery from disruptions, says Van den Akker.
Ground processes
The second PhD candidate, Lisanne Heuseveldt, is developing an algorithm that looks at ground processes, and then especially baggage handling and passenger boarding. Van den Akker: "Here too, you want to be able to quickly adjust your planning if an aircraft arrives too late or too early. We are now doing this for loading and unloading baggage and we also want to do this more in conjunction with other aspects, so we are looking at both baggage handlers and passenger boarding. Based on all kinds of considerations that the algorithm takes into account, you can then decide which flight has priority."
In addition to KLM, Utrecht University also works with NS, Prorail, Qbuzz and other organisations in the field of mobility and public transport in the AI & Mobility Lab.
Kickstart AI
Both projects are part of the Kickstart AI program. Within this program, companies such as KLM, Ahold Delhaize, ING and NS are working together on the development of AI applications.