How to optimise the airport queue?
How to optimise the airport queue?

Artificial Intelligence and the Airport - How artificial intelligence will save you time at the airport

Updated April 09, 2019 by CJ

Artificial intelligence is the next feasible wave of computing. The only thing holding it back is its unrealised value in practical applications. We are here today to provide some examples of how artificial intelligence can save you time at the airport.

There are many ways in which artificial intelligence can improve the airport experience, the focus of this post will be on queues. Recent research suggests that artificial intelligence can improve queues, meaning you will waste less of your life standing in line.

What is a queue?

A queue forms when there a greater number of actors requiring service than there are servers providing that service. Traditionally the branch of mathematics that tries to optimise a queue is called (funnily enough) queue theory.

The two predominantly frustrating queues at an airport are the check in counter and the security screening.

Using Artificial Intelligence

The basic process is:

  1. Collect data
  2. Develop model
  3. Train model with data
  4. Run model to generate predictions

Developing and training the model is a complicated process, one of the reasons artificial intelligence isn’t widespread. In this case though, there are already models to optimise queues such as these.

What is the problem here then? Data collection.

There’s an old computer science adage garbage in, garbage out. To generate good predictions, you need good data collected.

With good data, a good model (such as a neural network) can be built and trained and output collected. An appropriately built model will help the airport manage its staff to provide a better experience for customers.

A model built like this can find the right number of check in staff, security staff, counters, and screening facilities to have open and operating per hour of the day.

The idea is to pre-empt the needs of the passengers, rather than react when the queue grows to long (the domain of current queue management).

Why bother? (Where is the money?)

To add value, artificial intelligence must either increase revenue or decrease cost to the airport.

Cost, in these queue examples, is the deployment of staff and resources to serve passengers. Reduce staff numbers by removing idle staff and costs can be reduced. A predictive model of how to schedule staff in advance can assist in reducing this cost.

The better way to add value however is to allow the airport to process more passengers. More passengers mean more airline fees for the airport and a greater opportunity for shopping. Less time spent in queue means a better experience for passengers - passengers can be more likely to return, or to spend more time shopping.

Where next?

Artificial intelligence can do far more for an airport than replacing service staff with computer screens.

Artificial intelligence can work in the background to optimise processes and provide you with a more seamless experience. All we need is buy-in from the airport, and a good value proposition where the savings outweigh the costs.

Are you sick of standing in queues? Check out fun-sized airports for a great alternative.