Machi Zawidzki D.Sc. is presently a postdoctoral fellow (JSPS) at the Research Center for Advanced Science and Technology of the University of Tokyo. His research is titled "Improvements of the Seniors' Quality of Life through Application of Innovative Computational Systems to the Problems of Accessibility, Ergonomics and Housing & Living Environment." Prior to this, he completed the doctoral course (MEXT) at Ritsumeikan University with the dissertation "Application of Computational Intelligence to Engineering Design Problems in Architecture—Firmitatis, Utilitatis, Venustatis."
He loves elegance; he hates noise.
For portfolio and CV, see his website.
Project: Agent-Based Model (ABM) for Crowd Simulation
The main problem: exploration of the 2D CA rules in the context of traffic congestion.
- To find the behavioral patterns which result in the smoothest flow.
- Providing a reliable and flexible simulation method for any given environment.
Things to do:
- The speed of agents—it seems interesting to provide a model where agents can move at different paces, especially in the same flow.
- The size of the immediate proximity zone (perhaps 0 ~ 1.2 m—equivalent to the personal distance).
- The distance field.
- Measuring the flow: entering and leaving k agents.
- Subgroups of agents according to, e.g., age (different speed and/or CA rules).
- Sequence of intermediate goals (a path to be followed by agents in one stream).
- Define walls as adjustable polygons that can be dynamically changed even while running an experiment.
- Differentiate the agents visually (e.g. one can be traced).
Favorite Four-Color, Four State Turing Machine