Enhancing HiDAC Model Performance : Evaluating Unity Physics against Original approach in regards to diverse environments

Abstract: The High-Density Autonomous Crowds (HiDAC) model is an interesting crowd simulation model due to its unique ability to parameterize human characteristics. It is used in various applications such as urban planning, emergency management, and entertainment applications. Despite its widespread use, the HiDAC model faces limitations when dealing with large-scale, complex, and dynamic environments, impacting real-time applicability, accuracy in high-density crowd scenarios, and user experience. This thesis investigates the potential improvements in behavioral and computational performance of the HiDAC model by integrating Unity’s physics engine, a powerful tool for handling complex simulations with improved performance and scalability. The study focuses on the integration of Unity’s physics engine with the HiDAC model to improve the performance of collision avoidance in various 2D environments and compares it to the original HiDAC model using room-limited search. The results show that the HiDAC model integrated into Unity significantly improves collision avoidance and agent interaction in large, complex environments. By evaluating the simulation performance of both models in different scenarios, the research aims to provide valuable insights into the integration of Unity physics into existing crowd simulation models and explore its potential benefits in terms of performance, computational efficiency, and real-time applicability.