Modelling pedestrian navigation behaviour for crowd simulations

• Main Author: Yong, Alistair Jia Jin
• Other Authors: Cai Wentong
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Crowd simulation
• Online Access: https://hdl.handle.net/10356/181158
• Institution: Nanyang Technological University

Pedestrian navigation and crowd simulations have always been a useful tool across multiple sectors of society. Walking being almost a necessity when travelling, it is important to have accurate crowd simulation tools to predict and prevent possible disasters. With improving computational power and better technologies, we must keep trying to push the boundaries of what’s possible. For this report, we will be experimenting using Unity engine, which was originally created as a game engine, to see Unity’s capabilities in crowd simulations, along with incorporating AI through a relatively new Unity ML-Agents Toolkit to try and develop better crowd simulation tools. We concluded that Unity does have the potential to develop highly customized simulation tools to fit all sorts of purposes, and more time should be spent looking into the capabilities that Unity can provide.