ACED: Automatic Concourse Event Detection

Abstract

Fans of the San Antonio Spurs often face long delays when traversing the arena or waiting for food. Automatic Concourse Event Detection (ACED) is a novel system designed for tracking these statistics in the Spurs’ arena in real time. We use existing machine learning models and introduce novel processing algorithms to identify the total number of people in each section throughout the arena in addition to tracking the wait times for different restaurants and restrooms. ACED collects and stores this information in a database, which could be used to present fans with up-to-date arena information in a live dashboard to assist them in their in-game decision making. This would improve the overall fan experience, which could encourage fans to buy tickets more frequently. We provide the San Antonio Spurs with a completed implementation of ACED, which is ready to be deployed within the Spurs’ arena.