StreetSim V1.0

Using VR to simulate urban road-crossings

    • Date(s): September 2022 - January 2023
    • Topics: Virtual Reality (VR), Human-Computer Interaction (HCI), Immersion, Realism, Traffic Models, Crowd Models
    • Models: Intelligent Driver Model (IDM), A*/A-Star, Reciprocal Velocity Obstacles (RVO), Human Subjects Experimentation, Surveys & Questionnaires
    • Tech Stack: HTC Vive Pro, Unity (Game Engine), C#/CSharp
    • Collaborators:
    • Links: PDFGithubDOI

My initial goal for this endeavor was to create a virtual simulation of a urban road crossing, in its simplest form. While most urban road crossing simulations in research prioritize pre-determined, established gap distances between vehicles, vehicle velocities, or both, our study hopes to advance the practice by introducing high levels of verisimiltude with real-world scenarios by invoking unpredictability and stochastic modeling of phenomena in hyper-local interactions between pedestrians and vehicles. The main motivation is to see whether such stochastic, mathematical modeling would induce observers into conducting strategies and interactions typically seen in real-world scenarios. This is becoming ever more important as VR is poised to supplement real-world observational studies and offers the chance to observe human behavior in situations that are either hard to reach or too dangerous in the real world.

NoteMy first Ph.D. project!

This was my first major Ph.D. project. It was a massive undertaking, but I’m proud of what I accomplished with this. I managed to pour everything I knew about game development, VR, and user experiences to create this simulation. Much of the completely new things I had to first learn about are common traffic models used in simulations such as these, as well as how to make the vehicles behave realistically (or at least as close to realistic as possible). The pedestrians were much harder to emulate, but I think I got 75% of the way there. Modeling human behavior and its randomness is quite difficult…

NoteDownload the report here

Download the report here: Download Manuscript PDF (6.82 mB)

Gaze Behavior Visualization - Understanding Where People Look
NavMesh and Traffic Signal Crossing Logic
Inverse Kinematics: Replicating Head-Turning
Initial Exploration in Dynamic NavMeshes; Aberrant Agent Pathfinding
    • Kim, R., & Torrens, P. M. (2024). Building verisimilitude in vr with high-fidelity local action models: a demonstration supporting road-crossing experiments. Proceedings of the 38th ACM SIGSIM Conference on Principles of Advanced Discrete Simulation (pp. 119–130). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3615979.3656060, doi:10.1145/3615979.3656060
    • Full PaperBibtex