Spatial Assembly: Generative Architecture With Reinforcement Learning, Self Play and Tree Search
This work addresses the challenge of automated spatial assembly generation for architects, but it appears incremental as it builds on existing procedural generation and reinforcement learning techniques.
The authors tackled the problem of generating spatial assemblies in architecture design by combining reinforcement learning, self-play, and tree search with procedural generation algorithms, resulting in a method that learns policies to maximize designer objectives, though no concrete performance numbers are provided.
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.