Building MTR from Slime Mold Simulation

In this project, I delved into computational design using Grasshopper, mastering procedural modelling techniques such as data indexing, sorting, list manipulation, and mathematical surface generation. I began by experimenting with basic and advanced data structures to build responsive and parametric geometries, gaining a strong foundation in data-driven design logic.

Building on this, I explored urban simulation through agent-based modelling. Using the Physarealm plugin, I mimicked slime mold behavior to simulate natural pathfinding across the urban landscapes of Hong Kong and Kowloon. Due to limitations in Physarealm, I transitioned to Culebra, a more versatile plugin, to develop complex multi-agent behaviors including attraction, repulsion, flocking, and wandering.

To enhance realism and applicability, I integrated real-world data like population density, commercial hubs, and geographical obstacles. These inputs were used to simulate 18 overlapping runs, resulting in predictive mapping of pedestrian and transportation flow. I further extended the methodology to Dhaka, Bangladesh—proposing a data-informed, bio-inspired metro network in a city currently lacking full rail infrastructure.

This project demonstrated how natural systems like slime mold growth can be translated into computational tools for urban planning, offering a novel lens for designing efficient, organic transportation networks.