AI for Process Safety
Bridging the gap between frontier AI research and safety-critical industrial applications. Specializing in automated P&ID interpretation and deterministic engineering reasoning.
Frontier Research Areas
We focus on the most challenging aspects of industrial AI, where precision and reliability are non-negotiable.
P&ID Recognition
Advanced Vision-Language Models (VLMs) trained to identify equipment, symbols, and labels within dense engineering drawings.
Flow Path Analysis
Reasoning about connectivity and functional relationships—understanding what flows where and under what conditions.
Validation Boundaries
Developing formal benchmarking protocols and difficulty scales for AI in safety-critical domains.
Deterministic Reasoning
Moving beyond stochastic outputs to ensure AI systems produce repeatable, engineering-sound conclusions.
Digital Twin Integration
Closing the gap between static P&IDs and live operational data for next-generation digital twin systems.
Advancing Generative AI for P&IDs
Presented at the 22nd Global Congress on Process Safety (AIChE GCPS 2026), this work introduces the P&ID Question Difficulty Scale (PQDS v1), a binary evaluation framework for safety-critical AI, and a trimodal architecture fusing vision, graph, and language reasoning.
"Every connection, symbol, and note reflects an engineer's intent. The challenge is to make this information accessible to AI systems in a way that preserves that intent and enables contextual reasoning."
Advancing the Safety Frontier
Interested in collaborating on process safety research or implementing our P&ID analysis models?