eAI Automated P&ID Annotation: Try it in your Browser

We’re excited to announce that the hosted version of eAI – our AI-powered P&ID digitization tool – is now live!

Launch link → https://eai.plantfce.com

This web-based demo gives you a hands-on preview of our automated annotation pipeline, built to accelerate the extraction of piping, instrumentation, and equipment data from P&ID diagrams. No installation. No code. Just launch and annotate.

At this early stage, the demo supports shape detection and markup – the first step in a broader multi-stage annotation pipeline. Over the coming weeks, we’ll be rolling out the remaining steps to bring the full power of eAI to your browser.

What’s Working Now

• Standard Shape Detection: Recognizes key symbols such as pumps, vessels, valves, and instruments and marks them on the P&ID.
• Interactive UI: Adjust detection thresholds, preview markup overlays, and explore how the tool interprets the drawing.

Try the hosted demo: https://eai.plantfce.com

What’s Coming Next

We’re progressively bringing the full eAI pipeline to the hosted version. Here’s what’s in the works:

1. OpenAI Integration for OCR

We’ll integrate OpenAI’s models to extract tag numbers, equipment labels, and text fields from diagrams with high accuracy. [Coming soon: Implementation Details]

2. Walking the Lines

Our line tracing engine will follow process and instrument lines, determine connections, and structure them into a graph. [Coming soon: How We Build a Graph from a P&ID]

3. Inline Shape Detection

Using custom-trained models, we’ll detect small but critical inline components (e.g., flow meters, check valves, reducers) typically missed by template-based approaches. [Coming soon: Our Approach to Detecting Inline Components]

4. Export to a Graph Data Structure

The extracted data will be exported to a structured graph (e.g., Neo4j) for easy integration, validation, and querying. [Coming soon: From Drawing to Graph – Structuring P&ID Data]

5. Export to DXF

We’ll support re-export of annotated data to DXF format to integrate with CAD tools and SmartPlant P&ID workflows. [Coming soon: DXF Export for CAD Interoperability]

What We’re Holding Back

DEXPI Export

DEXPI export is currently on hold. The Python library is still in active development, and we’ll revisit this once it stabilizes for production use.

Stay Updated

This blog post will be continuously updated as new steps are added to the hosted version of eAI. We’ll also publish deep dives into each stage of the pipeline across:

• Storm Consulting Blog – For implementation insights, models, and engineering solutions
• eAI Learn Page – For technical documentation and usage guides
• PlantFCE Product Updates – For feature launches and case studies

Try the Demo, Share Your Thoughts

You can try the hosted version now at https://eai.plantfce.com

Feedback and feature requests are always welcome. Reach out via LinkedIn, email, or through our contact page.

Read More: Data Extraction from P&IDs Using eAI