AGIFORS Isn’t Trying To Sell You Anything

AGIFORS, the Airline Group of the International Federation of Operational Research Societies, has a rule that most industry events don’t: a technical paper cannot name a product, and vendors are not there to pitch their products. The rule applies to the peer-reviewed technical program specifically; sponsor content still exists at AGIFORS events, but it is fenced into its own clearly labeled slots, independent of the papers under review.

The group traces itself to 1960, when six Operations Research workers from Air France, British European Airways, KLM, Sabena, Swissair, and Trans Canada found themselves at the same international conference in Aix-en-Provence, France; they kept talking after the sessions ended. That conversation turned into a committee, and the committee organized a symposium at Spring Valley, New York, in October 1961. The symposium proved valuable, so they decided to keep going. AGIFORS exists because the content of the meetings created value for those involved, not because anyone set out to found an institution.

Research archives date back to the very first symposium and are worth a glance. The papers on record from 1961 cover crew rotation problems, telephone reservations staffing, engine reliability, and, plainly, “Comments on the Role of Operations Research in Airline Operations.” This foundational operational research work is what keeps airlines running, staying true to the group’s roots, regardless of the topic under discussion.

AGIFORS is a federation of working groups, and the Annual Symposium is in its 66th year. It is the group’s flagship meeting and will be held in Istanbul on October 25, 2026. Underneath this umbrella, Study Groups run on their own calendars throughout the year: Revenue Management and Distribution, Scheduling and Strategic Planning, Crew Management, Airline Operations, and Aircraft Maintenance. Each Study Group does the specialist technical work its own members need. The Revenue Management Study Group meeting is coming up soon, September 14-16 in Seattle, US, hosted by Alaska and Hawaiian Airlines. The meeting was originally planned to take place in Doha but was postponed and moved due to tensions in the Middle East. Thank you to the organizers and hosts who pivoted quickly to ensure the 2026 meeting could still happen.

It is interesting to look at the content from year to year and notice how the industry’s focus shifts over time. It is also fascinating to analyze the topics in light of a broader context that includes government regulations (a major one being the Airline Deregulation Act in the United States in 1978). As airlines went from mainly carrying mail and a few select passengers in a heavily regulated manner, to competing fiercely for market share and passenger loyalty, the folks at AGIFORS were trailblazing the theory, systems, and techniques that allow some of the most complicated logistics and operational challenges to be solvable in practice.

The airline industry has never had a shortage of problems to solve. In the 1980s, overbooking, nested seat allocation, EMSR, and demand forecasting became the core disciplines of leg-level revenue management. This foundation was the groundwork for network revenue management, which extended these concepts to the O&D itinerary level to solve the connecting-passenger valuation problem that leg-level control could not address. Into the 2000s the conversation shifted to more customer-centric themes like loyalty, bundles, and framing choice modeling as “treating your customers as human” (Rick Elieson, 2010). In this paper, regarding choice modeling, Elieson intriguingly drew a line back to Plato who pictured “the mind as a chariot. Emotions were like the horses, and logic and reason the charioteer.”

Then something interesting happened. In the 2010s, the ULCC business model began to gain momentum and the AGIFORS conversation shifted to ancillaries. Operationally, all of the prior science which enabled mainline carriers to optimize their entire network, rather than just individual legs, had to adjust. Researchers found new ways to innovate, enabling Network carriers to compete with the new breed of airline. This innovation gave rise to dynamic pricing, reinforcement learning, and the first hints of artificial intelligence—which brings us to the 2020s. As you may well imagine, more concentrated levels of AI are coming into focus, as exemplified by Kalyan Talluri’s fascinating 2025 presentation on training a model similarly to how LLMs are trained, “OMGPT: A Generative Pre-trained Transformer for Operational Decision Making”.

Running underneath all eras, present as early as that first 1961 symposium, is a question that has never fully settled: how much of the decision belongs to the machine, and how much belongs to the person watching it? In our next article, we will dive deeper into some of the key papers over the years that pose this question, including a few that will be presenting along this theme in 2026.

For over six decades, AGIFORS has attracted the most intellectually curious and academic minds inside the challenging world of airline operations. Year after year their members continue to stay true to the original purpose of the group. Simply put, they are dedicated to learning for the sake of knowledge and to sharing that knowledge for the greater good. It is something that is becoming less and less common these days, and subsequently more and more valuable.