Geometry-Based Swarm Grammars
06.03.2025Revolutionize an agent-based geometry generating approach! Suitable as a Bachelor's, Master's thesis or Game Research Lab.
Background
Developmental Biology has been called "the premier discipline" of science as it explains the miracle of life in minute detail considering all so many facettes of physical, chemical and biological (self-)organisation in the context of dynamical natural constraints. Accordingly, it seems a viable hypothesis that if we master modelling, simulation, optimisation and, eventually, implementation of complex synthetic organismal systems, we would as well be able to succeed in numerous other challenges of mastering complexity, e.g. in medicine, infrastructure or economy.
Swarm Grammars (SGs) are an agent-based approach to generating complex geometric structures considering spatiotemporal constraints and deploying numerous agent abilities inspired by biological organisms, e.g. near-field sensing, migration, reproduction,and differentiation, that are held together by reactive behaviour definitions. In this project, a novel SG approach is developed that focusses on the geometrically-driven growth of forms. In particular, virtual agents are placed on the many surfaces of an 3D object and extrude its corpus in different directions based on the agents' outwards accelerations. This SG approach will empower modellers to quickly retrace self-organised structural formations as observed in chemical, physical and biological systems.
Tasks
You are tasked to flesh out a geometry-based Swarm Grammar model, to show its effectiveness with respect to different modelling domains, e.g. plant growth and coral growth, and to offer according basic model instances. The base model and its instances should be presented geometrically and mathematically, the simulated outcomes should be visualised. The growth processes should further be impacted by changes in the environment, e.g. by interplay with other agents or obstacles, forces like gravity, wind, or the volatile availability of nutrients/construction materials. The scope of the considered factors and application domains can be adjusted to fit both a Bachelor' and Master's thesis.
Literature
[1] Mußmann, Marc, Daniel Nicolas Hofstadler, and Sebastian von Mammen. "An Agent-based, Interactive Simulation Model of Root Growth." Artificial Life Conference Proceedings 36. Vol. 2024. No. 1. One Rogers Street, Cambridge, MA 02142-1209, USA journals-info@ mit. edu: MIT Press, 2024.
[2] von Mammen, Sebastian, and Christian Jacob. "Genetic swarm grammar programming: Ecological breeding like a gardener." 2007 IEEE Congress on Evolutionary Computation. IEEE, 2007.
[3] von Mammen, Sebastian, et al. "Interactive simulations of biohybrid systems." Frontiers in Robotics and AI 4 (2017): 50.
