MATTER – Digital Material Database

MATTER is an experimental, digital material database. The collection includes conventional materials such as aluminum and carbon fibers, as well as biodegradable and organic substances, including chitosan, flax fibers, and mycelium. In addition to standard material categorization, the system features a network module. This module shows possible material combinations using generative, agentic tools. Materials are analyzed and algorithmically linked based on potential chemical, mechanical, and traditional connection types.

Objectives and Mitigation of AI Errors
The system does not claim scientific validity; instead, it serves designers as a tool for ideation and a starting point for research. To minimize hallucinations and systemic bias from the generative models, multiple models with interfaces to online sources are integrated into the core system. While this does not completely eliminate misinformation, it reduces its occurrence.

Technical Architecture and Open-Source Licensing
The technical foundation of the prototype is a flat-file database directly connected to the respective models via APIs. Additionally, a Wiki API feeds supplementary data into the system. MATTER is designed as a web-based, open-source solution. The application requires minimal hardware resources to ensure high system compatibility and can be modified or extended for custom requirements.

NETWORK Modul