AEC AI and Tech Strategy Podcast AI Augmented Drawing Review in AECO Industry – Ep 101
4 snips
Feb 4, 2026 
Luke Reeve, Principal Solutions Architect at Twin Knowledge with a background in structural engineering and applied AI. He explains how AI learns to read 2D drawings and PDFs. He highlights high-value use cases like repetitive review tasks, connecting dispersed information, and non-compliance detection. He contrasts augmented assistants with automated agents and offers practical starting points for firms.
AI Snips
Chapters
Transcript
Episode notes
Two Weeks Comparing Steel Drawings Sparked AI Work
- Luke Reeve started as a structural engineer and spent weeks manually comparing steel design drawings to fabricator shop drawings.
- That repetitive two-week review spurred him to study applied AI and join Twin Knowledge to automate those exact tasks.
Drawings Are A Visual Language For AI
- Drawing sets are a visual language AI must learn using computer vision plus multimodal and fine-tuned LLMs to extract title blocks, details, and tables.
- Consistent, structured data makes pattern recognition far easier; inconsistency increases model complexity and error risk.
Start With Repetitive Drawing Reviews
- Start AI pilots on highly repetitive, manual review tasks like submittal and shop drawing comparisons to get quick value.
- Use AI as a first-pass reviewer and keep an experienced engineer to validate edge cases and implicit knowledge.