Demand-side Incentives to Mitigate Wait Times at Airport Security Checkpoints
Airport security lines have become a major bottleneck of air travel operations, a major source of passenger dissatisfaction, and a recurring news story. The coupling of strong security requirements and high air traffic demand is such that long wait times routinely occur at busy airports worldwide. Record-high delays of 3 hours were reported during the busy Summer travel season in 2016 at some airports. The economic impact of these security delays is undoubtedly significant. First, they can lead to passengers missing their flights, and downstream disruptions across the travel industry. Second, in anticipation for security wait times, many passengers, encouraged by the airports and airlines, adjust their arrival times at the airport to ensure sufficient buffers. At the same time, airport security processes need to ensure compliance with strong security processes and requirements. Therefore, any operational or design change must ensure that existing risk levels will not be deteriorated.
On the supply side, the speed of screening processes could be increased through higher staffing levels and better screening technologies. These interventions, however, are limited by technological, institutional, and risk management constraints. On the demand side, the number of passengers screened everyday cannot be altered. However, the arrival patterns of passengers at the security checkpoints can be smoothed out over time and space to mitigate wait times given the requirements of the screening process. This is the focus of this project.
This project aims to design demand-side incentives to balance passenger demand over time and space. The core idea is to provide customer incentives to reduce demand loads at major bottlenecks by (i) measuring and predicting wait times at the different checkpoints, (ii) sharing such information with passengers, and (iii) designing non-monetary incentives to encourage passenger behaviors that will reduce demand overloads at peak hours. This can occur through three distinct mechanisms:
- Spatial smoothing: When a passenger arrives at the airport and is about to go through security, he/she can decide which checkpoint to go through. At PIT, for instance, the two alternative checkpoints often exhibit significantly different wait times. Sharing information on wait times can help balance passenger arrivals across the two checkpoints.
- Tactical delay: When a passenger arrives at the airport, he/she can voluntarily elect to delay the time he/she goes through security, if wait times are perceived to be high and projected to go down in the near future (assuming that the passenger has enough time before his/her flight).
- Strategic arrival: When a passenger plans his/her trip to the airport, he/she could strategically time his/her travel based on expected wait times at the security checkpoints.
In the first phase of the project, we will measure and predict wait times at different checkpoints in advance and in real-time. Data sources include published flight schedules and data provided by the airport on passenger behaviors. In particular, the airport is planning the deployment of a wireless technology to collect information on passengers’ travel patterns within the terminal buildings. This will be used to build, calibrate and validate prediction models based on data mining methods and queuing theory methods. Ultimately, this will identify the main bottlenecks of operations and the main opportunities for wait time mitigation. Note that the development of this first phase will be contingent on the quality of the data that will be available to us (e.g., wireless technology data from PIT).
In the second phase of the project, we will use these wait times predictions to inform the design of demand-side incentives to balance security queues in time and space. We will propose a field experiment design to test the effectiveness of such demand-side incentives in a real-world setting at PIT. Typical levers include: (i) information incentives by sharing real-time information on wait times, (ii) monetary incentives, which aim to use financial rewards to incentivize certain behaviors, (iii) social and behavioral incentives based on peer pressure and social influence, which aim to leverage potential influences within networks of friends or similar individuals, (iv) gamification-based incentives, which aim to use non-monetary rewards such as game scores or loyalty points to incentivize individual behavior, and (v) pro-social incentives, which aim to leverage the system-wide benefits of individual actions to incentivize certain behaviors for social good.
These interventions will be proposed to the airport in a field experiment design. If successful, this project can lead to a follow-up study to implement this field experiment at PIT, and assess the impact of the various interventions on passenger behaviors and security wait times.
This project will be organized in four phases:
1. Data collection (August – December 2018): Collection of operational and flight schedule data and wireless data from the airport. Descriptive analytics to gather insights on current wait times, and passenger behaviors in airport security processes.
2. Prediction model (September 2018 – March 2019): Development of prediction models to estimate security wait times, based on data mining and queuing-theoretic methods. The scope of this task will be contingent on the data made available by PIT.
3. Experimental design (January – May 2019): Development of a field experiment design that captures practical constraints and requirements, enables the collection of data throughout the field experiment, and thorough assessment of the effectiveness of the interventions.
4. Report writing (June 2019): The final deliverable will include (i) details on the prediction models, (ii) insights from predicted wait times and major wait times mitigation opportunities, (iii) description of proposed demand-side interventions for wait times mitigation, and (iv) a proposed field experiment design for future implementation.
Pittsburgh International Airport
Prof. Alexandre Jacquillat
Heinz College, Carnegie Mellon University
Prof. Beibei Li
Heinz College, Carnegie Mellon University