Creating digital P&ID diagrams has traditionally been a manual and time-intensive task. Engineers often spend hours sometimes days rebuilding existing drawings inside SCADA or HMI systems, carefully placing each component, reconnecting pipes, and re-entering tag information. The process is repetitive, detail-heavy, and difficult to scale across projects.
What is OD-Script?
OD-Script (Object Detection Script) is an intelligent automation tool that transforms static P&ID (Piping and Instrumentation Diagram) drawings into fully interactive HMI pages inside THMI.
Instead of manually recreating diagrams, users can simply upload a scan, photo, or PDF of an existing diagram. OD-Script automatically converts it into a structured, digital HMI layout ready to use and easy to edit.
What traditionally takes hours or even days of engineering work can now be completed in just minutes.
Why It Matters
When onboarding to THMI, one of the first requirements is a digital version of existing plant diagrams. In reality, these diagrams are often only available as:
- Scanned PDFs or photos
- Printed CAD drawings
- Low-quality images captured on-site
Rebuilding these manually is time-consuming, error-prone, and costly.
OD-Script eliminates this bottleneck by automating the entire conversion process—reducing effort, accelerating onboarding, and improving consistency.
How It Works
OD-Script processes each diagram through a series of intelligent steps, gradually transforming a static image into a structured, interactive system.
1. Equipment Recognition
The process begins by identifying all recognizable components within the diagram, such as sensors, valves, pumps, heat exchangers, and other equipment.
Using advanced AI models trained on industrial diagrams, the system detects and classifies each symbol while ensuring accuracy across both small and large drawings.
Result: A structured list of detected components, including their type and position.
2. Label Identification
Next, OD-Script reads the text labels associated with each component such as tag names and identifiers.
It uses optical text recognition to extract this information and intelligently matches it against available tag lists when provided. This helps ensure consistency and reduces interpretation errors.
Result: Each component is enriched with meaningful labels and tag references.
3. Connection Mapping
The system then analyzes how components are connected by tracing the lines that represent pipes and flows.
It reconstructs connections between elements, identifies intersections, and captures visual attributes such as line styles and colors where relevant.
Result: A complete map of how all components are interconnected.
4. Layout Optimization
To ensure clarity and usability, the extracted diagram is refined into a clean and consistent layout.
Elements are aligned, scaled, and adjusted to create a structured visual representation that is easy to read and interact with.
Result: A well-organized diagram with consistent positioning and proportions.
5. HMI Generation
Finally, the processed data is assembled into a digital format that can be directly used within THMI.
Each component is linked to standardized HMI elements, labels are preserved as metadata, and connections are visually represented. The result is a fully functional and editable HMI page.
Result: A ready-to-use interactive diagram that integrates seamlessly into THMI.
The Outcome
OD-Script bridges the gap between legacy engineering documentation and modern digital systems.
By automating the conversion of P&ID diagrams into interactive HMIs, it:
- Reduces manual engineering effort
- Speeds up project onboarding
- Improves consistency and accuracy
- Enables faster access to operational insights
Built for Real-World Conditions
OD-Script is designed to handle the realities of industrial environments—whether diagrams are old, scanned, or captured in less-than-ideal conditions.
It brings structure to unstructured data and turns static documentation into a dynamic part of your operational workflow.
At THMI, our ML specialist Marina Bortoli is continuously refining and enhancing our Object Detection feature. If you’re curious about the technical details behind it, feel free to reach out to her with your questions.
