Brussels Airport

Demand & capacity balancing tool with agent-based simulation

Who is Brussels Airport Company?

It is the company running Brussels Airport, the international airport located in Zaventem. More than 26 million passengers arrive or depart at Brussels Airport per year. It is also home to around 260 companies, together directly employing 20,000 people. It is a city in itself, bustling with activity, innovation and engagement.

The challenge

Ensuring a smooth welcome to passengers that are beginning their journey is a must that can be challenging in the ever-changing conditions of an airport. Long waiting lines and processing times are directly linked to passengers satisfaction. However, waiting lines dynamics are very complex complex and cannot be analysed with simple mathematics.

Our approach

To optimise the planning of resources and keep the waiting times under control, Jetpack built a series of simulations tools to optimise the planning of resources such as check-in equipments (counters, kiosks, bag tags & drops) or security screenings resource planning :

  • Multi-agent models to simulate passenger arrivals, queuing and processing at the desks, using simulations to capture real-life behaviour variations.

  • An online app to parametrize the model and visualize simulation results (queue lengths and passenger waiting times throughout the day). The airport was able to assess the impact of their planned desk allocation on passenger experience.

Tech stack

The whole solution is hosted on Amazon Web Services (AWS). The simulation engine consists of a series of python packages built using Docker and stored on AWS Elastic Container Registry (ECR). The simulation engine is executed on demand from the front-end application on AWS Lambda.

  • Outputs of the simulation are written in a PostgreSQL database hosted on Amazon RDS (Aurora).

  • The API is built using nodeJS and serverless and hosted on API Gateway and AWS Lambda.

  • The front-end web application is built using ReactJS. Charts and visualizations combine both d3.js and Plot by Observable. The application is hosted on AWS S3 and served using AWS Cloudfront.