Spatial Assembly: Generative Architecture With Reinforcement Learning, Self Play and Tree Search

TL;DR

This paper introduces a novel method combining Reinforcement Learning, Self-Play, and Tree Search to generate spatial assemblies for architecture design, inspired by procedural algorithms like WFC, to optimize design objectives.

Contribution

It presents a new approach that integrates RL and procedural generation techniques for spatial assembly creation, enabling optimized architectural designs.

Findings

Successfully generated spatial assemblies satisfying design constraints.

Demonstrated application in architectural design scenarios.

Achieved optimization of design objectives through learned policies.

Abstract

With this work, we investigate the use of Reinforcement Learning (RL) for the generation of spatial assemblies, by combining ideas from Procedural Generation algorithms (Wave Function Collapse algorithm (WFC)) and RL for Game Solving. WFC is a Generative Design algorithm, inspired by Constraint Solving. In WFC, one defines a set of tiles/blocks and constraints and the algorithm generates an assembly that satisfies these constraints. Casting the problem of generation of spatial assemblies as a Markov Decision Process whose states transitions are defined by WFC, we propose an algorithm that uses Reinforcement Learning and Self-Play to learn a policy that generates assemblies that maximize objectives set by the designer. Finally, we demonstrate the use of our Spatial Assembly algorithm in Architecture Design.