Description:

FRAMECAD (https://www.framecad.com/) is the world-leading cold formed steel (CFS) construction automation solution provider that exports end-to-end design and building systems to more than 35 countries. With FRAMECAD’s technology, a building either residential or commercial can be produced and assembled into flat packs in off-site factories and then delivered to construction sites for installation.

 

This project aims to develop steel frame quality checking technologies using computer vision techniques. Quality assurance of framing provides confidence that the frames have been assembled in the manner in which they were designed and engineered. Inaccuracies can be addressed in the offsite location rather than be identified and resolved onsite.

 

Current process

Cold-formed steel frames are assembled in factories before being delivered to construction sites. The current quality-checking process relies on inconsistent ad-hoc checks from human operators.

 

Proposed solution

Use computer vision technologies to automatically complete QA checks of steel frames and generate standardized QA reports before frames leave the factory.

 

• Evaluate the feasibility of computer vision technologies for identifying quality issues with steel frame assembly and report on its algorithm performance limitations. Key functional features include:
  • Based on a design drawing, locate and verify the placement of screws in a frame. This may occur once all assembly is completed or may instead be a function call once assembly is created.
  • The QA function should identify and report quality issues based on a discrete set of criteria (e.g. missing screw, incorrect screw pattern, incorrect screw type, incorrect application (screw not all the way in)).
• Understand computer vision technologies' implementation opportunities and limitations in such application contexts.
• Develop a prototype system to validate QA workflow and generate QA reports.
• Extension: Integrate the output of the QA function into the FRAMECAD NEXA tool. This way, a QA report will be available for each frame.

Type:

Undergraduate

Outcome:

Prerequisites

None

Specialisations

• Mechatronics Engineering

Categories

• Industry 4.0 Smart Manufacturing Systems
• Mechatronics

Supervisor

• Yuqian Lu

Team

• Cole Parker
• Chi-Hsiang Yang

Lab

Manufacturing Systems (405.870, Lab)