Gravity model for air passenger demand estimation: the addition of big data

Gravity model for air passenger demand estimation: the addition of big data

Boelrijk, Wesley (TU Delft Technology, Policy and Management)

Annema, Jan Anne (mentor)
Enserink, Bert (graduation committee)

Delft University of Technology

Engineering and Policy Analysis

2019-08-09

Over the last decades, flying has become increasingly accessible and the global aviation industry is growing rapidly. This causes airports and airspaces to reach their capacity limits. Reliable estimation and forecasting of air passenger demand is very important for airlines in order to allocate resources effectively in a constrained and competitive aviation environment. A model is developed to estimate air passenger demand suitable for routes where currently no direct service exists in order to aid decision-making on new destinations. For this model the traditional gravity model is extended with big data from worldwide flight search engines (meta search data). The model is applied to a case study with real world data from KLM Royal Dutch Airlines. We find that the gravity model enriched with meta search data is able to accurately predict demand and rank destinations that are currently not connected to KLM’s network by a direct flight. The data-driven approach allows us to give recommendations about promising destinations that could be added to an airline’s network. Ultimately, a more efficient network as a result of advanced demand estimation is expected to lead to higher load factors and lower emissions per passenger.

Air Passenger Demand Estimation
Aviation
Networks
Gravity Model
Ordinary Least Squares
Big Data

http://resolver.tudelft.nl/uuid:2f2ee8aa-3ed3-4e5d-802d-b64e85853d7a

Student theses

master thesis

© 2019 Wesley Boelrijk