A Deep Learning Method for Detecting Defaults on Assembly Line

With the advancement of methods based on artificial-intelligence, computer vision and deep learning, activities concerning progress monitoring, safety management, and quality control can be automated that leads to saving in time and cost. With the collaboration with partner industry, computer-vision approaches are going to be employed in order to find an improved method that utilizes new machine learning techniques to detect defaults on prefabricated production line. The results will help with improving the quality level of products and reducing potential project risks. The area is relatively young; especially for construction industry; and a lot of R&D works need to be done in order to increase the efficiency of the current procedures. As the market for AI-based quality monitoring tools is rapidly growing, the results of this study will greatly benefit construction industry by offering a possible improved procedure to detect quality issues in production line.

Faculty Supervisor:

Ali Motamedi;Daniel Forgues

Student:

Partner:

Groupe Canam

Discipline:

Engineering

Sector:

Construction and infrastructure; Manufacturing

University:

École de technologie supérieure

Program: