Description
Architects and BIM teams are generating AI images faster than ever. But outputs rarely connect back to real materials, specifications, or fabrication. This session will draw on three MIT Architecture and Computer Science and Artificial Intelligence Laboratory (CSAIL) research projects to show exactly where the design-to-build pipeline breaks down and what fixes it. What we'll cover includes steering AI image generation with element-level references, matching rendered materials back into BIM specifications, and connecting natural language to robotic assembly each finding maps to a workflow that practitioners can act on. You'll leave with research-backed behavioral changes where AI tools genuinely add value in their pipelines and where the gaps need bridging.
Key Learnings
- Discover where AI outputs break down in the design-to-build pipeline.
- Learn how to match AI-generated renders to real material and sustainability specifications.
- Learn how to connect generative design outputs to physical fabrication using vision-language models.
- Learn where AI adds real value in AEC and where it falls short backed by research at MIT.