Airports are a catalysator of economic activity, economic growth, and welfare in a region. Boosting airport activities only strengthen this effect. However, airport activities cannot just expand. Many major airports, most notably Brussels Airport and Schiphol Airport, are bound by spatial limitations and accessibility limitations, for example traffic congestion, public and transport connections.
Boosting airport activities hence must come also from within, by optimizing the airport operations end-to-end. With two goals in mind:
Passenger experience: The passenger ought to receive a positive experience from the moment the passenger enters the airport until the moment he leaves. This means maximizing the time spent on pleasant activities (e.g. taking a coffee or shopping) and minimizing the time lost by waiting for check-in, registration, customs, and luggage.
Benefit from the passenger’s time: During the time the passenger spends in the coffee-bar, bookstores, and shops, he initiates economic activities to the benefit of the airport and its partners. The time the passenger loses in queues is not only time lost. It is also a symptom of congestions requiring more staff and hence more expenses to resolve them.
However, airport operations are organized in silos. Processes are steered by different units with a highly focused span of attention. An all-encompassing view that could show how different processes either refrain or help one another is not available.
The airport eco-system
An airport is an economic entity, but also an enabler for economic activities of many different players. The airport eco-system includes, besides the airport itself, also the airlines, air cargo service providers, security providers, catering companies, flight control, technical maintenance providers (MRO), Customs. Not to forget the customers: the passengers and air-cargo shippers.
Optimizing airport operations touches upon all of those.
The planning challenge
All of the parties in the airport eco-system are confronted with – to a higher or lesser degree – the same planning challenges:
Seasonal variation: The volume of airport activities follows seasonal tendences. This is most intuitively so for passenger transport, but also air freight has peak periods (typically 4th quarter).
Daily irregular demand curves: The volume of activities evolves drastically also with the course of one day. Brussels Airport for example has a high peak of inbound and outbound flights in the morning, and one in the late afternoon. Most airport activities are directly linked to arriving and departing flights and have hence to cope with this daily irregular pattern. They must assure that they can staff all their operations on the peaks and avoid ‘idle time’ (having too many people present during the hours where there is almost nothing to do).
Air traffic disturbances: Flight arrival times are hard to predict. Flights can arrive early or late. Departure times are in principle better under control. Nevertheless, late departures are also not exceptional. All these disturbances, small or large, continuously modify the flow of activities within the airport premises.
Taking a global picture
All parties in the airport eco-system handle the planning challenge in their own way, starting from the flight schedule. They try to optimize the allocation of their resources, unaware of how the other parties on the airport ground are handling the same challenge. Large queues at the check-in implies that a group of passengers is going to be late and hurrying through the airport buildings. Security checks that have not foreseen this may jeopardize the departure time of the flight, possibly also generate waiting lines on the departure runway.
An optimal airport flow requires that all parties responsible for part of the airport activities take planning decisions that are in-line.
Flexibility to respond ‘now’
Decent up-front planning is only part of the answer. It is a utopia that all bottlenecks can be smoothened by planning carefully up-front. The combined activity of all passengers in the airport building is a complex dynamic system. Tens of unforeseen events that occur during the day – a plane arriving late, a technical failure, a bottleneck in the underground, a large family with passport issues – continuously modify the flow.
The ambition to optimize this complex dynamic flow requires:
The ability to understand the situation now, at any moment in time.
The ability to be able to react promptly, e.g., by moving a team of security agents from one terminal to another.
The ability to understand the effect in the very short term of those decisions further down the line.
What does this mean for medium- and short-term planning?
The medium-term (weeks ahead) and short-term (days ahead) plan remains the main vehicle for managing the operations efficiently. The flight schedule remains, as it is today, the main source of information to make those plans. An additional source of information can be insight in possible disturbances and the resources required to handle those disturbances optimally. Without any doubt, experienced planners have those insights and apply them to improve the stability of the plan. Today, most airports have the data allowing to formalize those insights and making them an explicit source of information to improve medium- and short-term planning.
What does this mean for immediate planning decisions (time zero)?
Coordination is key. There will always be multiple parties responsible for parts of the chain. Each of those parties expects to remain in control of the assignment of its assets. A central controller that is steering the entire airport would be optimal but not achievable. Through close coordination a near to optimal management process can be envisioned. It requires a centralized and shared technology, provided by the airport. This gives all actors an end-to-end view on current and very short-term airport operations, sharing and taking into account the operational decisions taken at different places in the chain.
Data and predictions
Today, airports dispose of data reflecting past and ongoing operations, e.g., data about passenger volumes in various airport locations, number of people that have already checked-in for a given flight. There is also data existing that could be exploited if it were shared pro-actively: i.e., data about number of passengers on a flight, pieces of luggage, number of business travelers versus tourist travelers, etc. The aviation sector is amongst the most digitized sectors. Information exits and, if it were better shared, can be made available to get a complete picture of ongoing and upcoming airport activities.
Machine learning techniques enable to uncover patterns and to better understand expected airport loads. Airports start to use them to make better passenger forecasts to the benefit of the medium-term plan (impacted by seasonal effects) and short-term plan (impacted by short-term effects such as the weather forecast).
The next logical step is to try to understand immediate effects (minutes up till one hour effect). The period between now and one hour from now is the period that you can still do something with the passenger experience for the passengers that are in the building now.