CoasterQueue - Tracking wait times with mobile phones

Abstract

Most of the time spent in a theme park is waiting in lines. In most parks, there are only two ways to determine the wait time of a ride. One is by physically walking to the ride entrance, where an estimate is posted. This can waste valuable time only to nd out that a ride has an incredibly long wait time. The other is by using an existing crowdsourcing application. However, these applications require active user participation to collect data, meaning the users must input the data manually. This can result in a lack of data and poor reviews. In this paper, we present a wait time estimation system for theme parks. This system is based on data collected by a mobile phone's sensors including accelerometer and rotational data. Using this sensor data, we detect movement patterns which determines the user's current state. We show that the patterns generated by rides, walking, and waiting are distinct enough to determine what the user is doing. Using sensor data, in conjunction with time, we are able to determine how long a user waited in line and which coaster they were waiting for. Using this method, we have developed a passive line tracking system which does not require any user interaction and can be used in any theme park in the world. We comprehensively test our system and the evaluation results show that the system achieves outstanding accuracy for wait times and classifying coasters.