Feature Extraction from Engineering Drawings with AI
Powered by AI, Werk24's API reads Engineering Drawings and returns a JSON response with structured information, including Title Blocks (Drawing ID, Designation, Material, General Tolerance), Measures, Tolerances, Thread, Chamfer, GD&T, Radii and Surface Roughness.
Cloud service through APIs
Simply submit your drawings to our API and within seconds get result as JSON format or Excel sheet.
This is the faster, easier way to integrate WERK24’s AI capabilities into your work processes or customer-facing products. It also scale flexibly with ever-expanding API features from Werk24.
WERK24 Products
ADDITIVE ANALYSIS
Identify parts for Additive Manufacturing
Unsure which parts are suitable for Additive Manufacturing?
Werk24 automatically assesses the suitability of Drawings with regards to a range of Additive Technologies
✅ Extract Outer Dimensions (often a limiting factor)
✅ Extract tight tolerances (often require a subsequent milling step or favor CNC)
✅ Prioritize parts based on AM feasibilities of the chosen AM methods
✅ Suggest alternative AM methods or extra milling steps
FEASIBILITY ANALYSIS
Automatically filter in-coming RFQs
Receiving RFQs that don’t match your machines?
Our AI reads Drawings and checks whether the parts match your manufacturing constraints.
✅ Extract production limiting factors incl., smallest Tolerances, smallest Thread diameter, smallest Radius, smallest GD&T
✅ Integrate customized filtering logic
✅ Convert a wide range of steel standards to EN10027 material codes and filter against a list of supported material
PMI EXTRACTION
Detect critical factors for production or pricing
Your online quoting system requires manual entry of PMI from technical drawings?
Werk24 extracts all relevant PMI.
✅ Material
✅ General Tolerances
✅ Tolerances & Fits
✅ Threads
✅ GD&T
Technical capabilities
Werk24's TechRead API utilizes Artificial Intelligence and Computer Vision to extract key information locked in scanned Technical Drawings. Unlike traditional OCR solutions, our technology can understand great amount of contextual information and convert it into structured and searchable digital repositories.
Access Title Block
Werk24’s API reads the non-standardized Title Block and translates its content into a structured machine-readable format. Unlike OCR, Werk24’s AI interprets the content even when captions are missing or located in another cell.
Extract Measures
Werk24’s AI extracts Measures and converts them into a standard data structure, allowing direct access to Nominal Size, Tolerance (Fit & Deviation), Thread and Chamfer information. It’s way beyond traditional OCR solutions that only returns unstructured Strings.
Capture GD&Ts
Our API captures the full range of Geometric Dimensioning and Toleration (GD&T) frames on drawings and returns them in a standard data model. You can directly access the characteristic, toleration and reference data.
Create Thumbnails
Separating Sectional Views can be challenging, especially when they are positioned very close to one another. TechRead API separates them and generates thumbnails for individual Sectionals, as well as the whole drawing.
Why choose us
First-of-its-kind technology
Leveraging Artificial Intelligence (AI) and Machine Learning (ML) technology, Werk24 provides industry-leading accuracy and intelligent enrichment for data extraction from Technical Drawings. With many advantages over generic OCR, our technology leverages context information and converts information into structured and CPU accessible JSON formats, which then become ready for your application to use.
High accuracy and continuous learning
Leveraging context information allows us to achieve high levels of accuracy. And when something is misread, it’s not for long. Our self-learning Artificial Intelligence learns from the processed data and improves its quality constantly.
Faster and streamlined processing
With Werk24 your application easily understands Technical Drawings - even when they were drawn 30 years ago and scanned multiple times. We process a wide range of formats (JPEG, JP2, BMP, PNG, TIFF, PDF), and automatically extract, understand and deliver the results within seconds - 24 hours a day.