AI-Enhanced Tool Detection in IFC Import Manager

The latest update to the IFC Import Manager introduces an experimental AI-driven function to detect tools. By leveraging AI, this feature expands the range of tools that can be detected during import, automating recognition for a variety of tools and simplifying the import process.

Supported Tool Types

Currently, the AI functionality can detect and import a limited range of tool types:

• Dove (male and female)
• Tenon (male and female)
• Drill
• Cone
• X-Fix

These tools are identified by the AI with a confidence score, enabling automated handling of tool types that were previously undetectable within the IFC Import Manager.

Confidence-Based Detection and Visualization

The AI evaluates each detected tool and assigns it a confidence score:

• Predicted Tools: Tools detected with a certain confidence level are marked in green for easy identification.
• Accepted Tools: Once a tool is confirmed—either automatically by the system or manually by the user—it is displayed in blue.

This visual cue allows users to quickly see which tools have been identified and validated, enhancing the ease and accuracy of tool management.

Manual Adjustment and Refinement

For greater control, users can manually refine the AI’s predictions. By using the Shift key and the lasso tool, it’s possible to select specific “fitter” tools within the model and adjust the tool detection results to better match the intended operation. This ensures the AI suggestions align closely with project needs.

Importing Tools with Complete Information

Once tools have been detected and accepted, they are imported into the model with the complete set of tool information. This includes relevant details such as type, configuration, and orientation, supporting accurate model representation and export.

‍