DETECTA
Ai-Based CAD Drawings
Ai-Based CAD Drawings
Assignment
1. st Semester, 2022-2023, Design Driven Project
Institution: RWTH Aachen University
Construction and Robotics Master
Project Team:
B.Sc. Arch. Hüseyin Alan
Task: Prototype Design & Development
B.Sc. Arch. Hossam Khaled
Task: Plugin Development
Tutor: M.Sc. Luca Fahrendholz
Individualized Production | RWTH Aachen
Director of CR-Master: Univ-. Prof. Dr. Sigrid Brell-Cokcan
Individualized Production | RWTH Aachen
Description
Detecta revolutionizes architectural surveying by using AI to automate CAD drawings, enhancing accuracy and efficiency. It integrates Meta AI's SAM with Autodesk's tools, streamlining the conversion of images into precise technical drafts. Reducing manual effort, this solution promises significant advancements in the Architecture, Engineering, and Construction industry.
Detecta revolutionizes architectural surveying by using AI to automate CAD drawings, enhancing accuracy and efficiency. It integrates Meta AI's SAM with Autodesk's tools, streamlining the conversion of images into precise technical drafts. Reducing manual effort, this solution promises significant advancements in the Architecture, Engineering, and Construction industry.
Introduction
Detecta addresses inefficiencies in architectural surveying documentation and technical drawing, crucial for the Architecture, Engineering, and Construction (AEC) industry. Traditional methods, often labor-intensive and error-prone, can lead to costly inaccuracies. This project is designed to utlise digital processes to minimise errors and increase project efficiency.
Detecta addresses inefficiencies in architectural surveying documentation and technical drawing, crucial for the Architecture, Engineering, and Construction (AEC) industry. Traditional methods, often labor-intensive and error-prone, can lead to costly inaccuracies. This project is designed to utlise digital processes to minimise errors and increase project efficiency.
Development
The project uses Artificial Intelligence (AI), particularly Meta AI's Segment Anything Model (SAM) and Autodesk's suite of tools (including AutoCAD, Revit, and Autodesk Construction Cloud), to automate the generation of survey and technical drawings. The workflow involves image definition, segmentation, edge detection, and CAD drawing creation, culminating in Revit 2D views and collaboration via Autodesk Construction Cloud. Detecta is a desktop plugin for Revit, chosen for its computational capacity and integration with Autodesk solutions.
The project uses Artificial Intelligence (AI), particularly Meta AI's Segment Anything Model (SAM) and Autodesk's suite of tools (including AutoCAD, Revit, and Autodesk Construction Cloud), to automate the generation of survey and technical drawings. The workflow involves image definition, segmentation, edge detection, and CAD drawing creation, culminating in Revit 2D views and collaboration via Autodesk Construction Cloud. Detecta is a desktop plugin for Revit, chosen for its computational capacity and integration with Autodesk solutions.
Application Output
Detecta significantly reduces the time required for creating technical drawings and enhances precision by eliminating common human errors. It fosters improved team collaboration and provides a central hub for project updates. The current limitation is the approximate nature of drawing scales, which the team plans to address in future versions.
Detecta significantly reduces the time required for creating technical drawings and enhances precision by eliminating common human errors. It fosters improved team collaboration and provides a central hub for project updates. The current limitation is the approximate nature of drawing scales, which the team plans to address in future versions.
Future Development
Plans for future development include enhancing the application with features like an AI line editor for refining CAD lines, a smart filter for tailored object detection, and a multi-projection feature to integrate point cloud data. These advancements aim to automate multiple CAD projection generations and increase design flexibility.
Plans for future development include enhancing the application with features like an AI line editor for refining CAD lines, a smart filter for tailored object detection, and a multi-projection feature to integrate point cloud data. These advancements aim to automate multiple CAD projection generations and increase design flexibility.
Conclusion
Detecta represents a transformative step in the AEC sector, efficiently converting photos into CAD drafts. It minimizes manual intervention, reduces costs, and democratizes architectural drafting. The team is enthusiastic about further augmenting Detecta with new features, indicating a promising trajectory for AI in architectural innovation.
Detecta represents a transformative step in the AEC sector, efficiently converting photos into CAD drafts. It minimizes manual intervention, reduces costs, and democratizes architectural drafting. The team is enthusiastic about further augmenting Detecta with new features, indicating a promising trajectory for AI in architectural innovation.
Acknowledgments
This project, "Detecta," developed during the Design Driven Project course at RWTH Aachen University's Construction and Robotics Master Program, incorporates cutting-edge tools and models to revolutionize CAD drawing processes in the Architecture, Engineering, and Construction industry. Our sincere gratitude goes to Autodesk for their comprehensive suite of tools, including AutoCAD, Revit, and Autodesk Construction Cloud, which played a pivotal role in the development and execution of this project. Additionally, the integration of Meta AI's Segment Anything Model (SAM), as detailed in the publication by Kirillov et al., 2023, has been instrumental in enhancing the AI capabilities of Detecta. The combination of these advanced resources has enabled us to push the boundaries of technological innovation in architectural surveying.
