Engineering Data & Automation
DWRS - Engineering Drawing and Catalogue Automation
A web tool that reads technical drawings and helps engineers turn them into product catalogues.
Illustrative system view
Context
Engineers needed to turn information from technical drawings into a usable product catalogue.
Challenge
Reading drawings, checking revisions and copying product details into a catalogue took time and required careful review.
Contribution
We built the web application and the document-processing tools, including OCR, data extraction, product grouping and Excel exports.
- PDF/TIFF processing
- OCR
- catalogue generation
- engineering review
- data validation
- visual reporting
- Django
- Python
Solution
The web application lets users upload drawings, review the extracted information and download a catalogue. It also provides visual summaries and logs for each analysis.
Drawing upload and processing
Built a web workflow for uploading PDF and TIFF drawings, launching background analysis, tracking progress, and preserving run logs.
Engineering extraction
Combined document parsing, OCR fallback, rule-based extraction, and validation checks to prepare structured technical information for review.
Catalogue output
Generated catalogue-ready Excel outputs while keeping uncertain, missing, or low-confidence fields visible to engineering users.
Similarity and clustering
Applied local machine-learning techniques to support similarity grouping, product-family review, and repeated catalogue preparation.
Visual review reports
Created visual PDF and chart outputs that help engineers inspect product families, similarities, missing fields, uncertain values, and review evidence.
How the system connects
- PDF/TIFF drawings
- OCR and parsing
- Extraction and validation
- Review workflow
- Catalogue and visual reports
Engineering considerations
- engineering judgement remains responsible for final review
- unclear or incomplete values are flagged instead of invented
- review evidence links extracted data back to source material
- confidential drawings, templates and product details are not published
Outcomes
- Created a maintainable base for turning technical drawings into structured catalogue data.
- Improved traceability from source drawings to extracted values, review evidence, and generated outputs.
- Reduced repeated manual preparation while keeping uncertain or incomplete information visible for engineering review.
Attribution & confidentiality
Developed by Artava Systems for Wabtec.
Drawings, customer data and other confidential project material are not shown.
LET’S FIND A WAY FORWARD
Have a project in mind?
Tell us what you need. We can start with a short conversation.