Quotation Factory

Feature Detection and Engineering Intelligence

Beyond global shape recognition, the system was extended to detect detailed manufacturing features critical for cost estimation and production planning. These included:

Bending operations
Holes and cutouts
Engravings and markings
Louvers and ventilation patterns
Flanges and connection features

Detecting these features required fine-grained analysis of local surface properties, topology changes, and geometric discontinuities within the CAD model. The combination of mathematical modeling and AI-based classification allowed the system to reliably interpret manufacturing intent from raw geometry.

Manufacturing-Oriented Computation

In addition to recognition tasks, the system incorporated several advanced manufacturing optimization modules, including:

Optimal material utilization and nesting strategies
Sheet metal unfolding and flattening algorithms
Toolpath generation for CNC manufacturing processes

These components ensured that the system did not only understand geometry but also translated it into efficient and production-ready manufacturing instructions.

Technology Stack

The system was built using a combination of high-performance geometry processing and AI technologies, including:

devDept Eyeshot (3D CAD visualization and processing library)
C++ and Python
Mathematical and geometric modeling techniques
Machine learning and artificial intelligence methods
Custom algorithms for feature extraction and classification

Impact

This system enabled a fully automated quotation pipeline where CAD files could be analyzed, interpreted, and converted into manufacturing cost estimates with minimal human intervention. By combining geometric reasoning with machine learning, the solution significantly improved speed, consistency, and scalability in the quotation process.