What is Discrete aggregation

Discrete aggregation is one of the hottest approaches of computational design, along with the emergent swarm intelligence and AI. It recognizes small distinct parts that can be assembled into larger aggregates according to certain requirements. The parts are often represented by material components assembled using connectors. Most prominent discrete approaches are voxelization, discrete growth and constraint modeling, such as Wave Function Collapse.

Discrete approach benefits from breaking down the problem into small parts. The parts are often represented by real physical components with real physical connectors. That allows to combine prefabrication with computational design, which is approachable with the current conventional technology. Discrete aggregates also scale well, so it is easy to prototype a small sculpture, scale up to a pavilion and eventually produce a large installation.

What is Wave Function Collapse

Monoceros is an implementation of the Wave Function Collapse (WFC) algorithm developed for game design by Maxim Gumin and extended and promoted by Oskar Stålberg with his game Townscaper.

WFC is an algorithm that fills an entire spatial envelope with Modules in a meaningful way. The envelope is divided into discrete box-like Slots. The algorithm selects and places Modules into the Slots, making sure any two placed Modules can become neighbors (do have a common connector) according to specified Rules.

WFC is not a growth algorithm – it always generates a complete valid aggregate (or no solution at all), leaving no Slots empty. It has important implications for design, architecture and urban planning: the aggregates are not branching, but create rhizomatic structures, there is no overlapping, there are no untreated areas.

It also requires a certain sensitivity and usage strategies, which makes WFC and Monoceros a methodology rather than just a tool. It can be used for projects of various scales and types and the use can be partial and shallow as well as thorough and informed. It is a great approach to materialize big data distributed in space into meaningful volumetric assemblies.