Crowd Flow Prediction Using Movement Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to accurately predict and manage movement flows of people at large events or amusement parks, making it difficult to estimate waiting times and control crowds effectively.
Innovation Solution
A method involving the spatial definition of locations, creation of movement graphs, assignment of usage probabilities and travel times, and dynamic determination of these factors using monitoring technologies like video, lidar, and near-field queries to forecast future crowd numbers and adjust accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If movement flows of people are to be predicted accurately, then monitoring technologies and data collection must be enhanced, but system complexity and cost increase
Solution Approach 1:
The system employs multi-functional monitoring technologies that can detect multiple types of data (location, queue length, waiting time) using the same hardware infrastructure. Mobile devices serve both as communication tools and as tracking instruments, while monitoring systems simultaneously capture spatial position and queue status, reducing the need for separate specialized devices for each measurement function.
Solution Approach 2:
The system utilizes data that users already generate through their own devices and behaviors. Mobile devices automatically provide location data through GPS and network triangulation without requiring additional sensors. Queue monitoring leverages the natural presence of users in the monitored area, converting passive user behavior into active data collection for prediction accuracy.
2Productivity
If real-time monitoring and dynamic adjustment of movement flows is implemented, then crowd management effectiveness improves, but computational resources and processing time increase
Solution Approach 1:
The system pre-calculates prediction models and movement patterns based on historical data and current conditions, preparing forecast information before it is actually needed for decision-making. By continuously updating prediction models in advance, the system reduces the computational burden during real-time crowd management events, as the heavy lifting of pattern recognition and forecast generation has already been performed.
3Measurement precision
If multiple monitoring technologies are deployed to track user movement, then data accuracy improves, but user privacy concerns and data security requirements increase
Solution Approach 1:
The system extracts only the essential information needed for crowd management from the data collected by monitoring technologies. Instead of tracking and storing complete user profiles or detailed movement histories, the system extracts aggregate location data, queue lengths, and waiting times that are sufficient for prediction purposes while minimizing personally identifiable information retention.
Data Source
Figure 1a~1b
Figure 2
Figure 3
AI summary
Methods for determining the movement of persons comprising defining at least three spatially separated locations for the persons, determining movement graphs of the persons' movements between any two of the locations, recording the number of persons at each location, and predicting the number of persons at each location after a certain time using the movement graphs.