Facade Measurement Nelissen
AI & Data
Semester programme:Artificial Intelligence
Client company:Nelissen BV
Aleksandar Konstantinov
Alpay Demirci
Juan Alejandro Sola Castermans
Project description
Can artificial intelligence offer a more efficient and accurate method for measuring the walls of a building facade compared to traditional techniques?
Context
Nelissen receives facade drawings from architects in PDF format. Based on these, the company produces an enhanced version that includes technical details related to heating, ventilation, and electricity.
However, the current process for measuring facade walls is outdated. Employees manually extract measurements using a ruler, which is not only time-consuming but also prone to human error.
The aim of this project is to develop an AI-powered tool that enables employees to upload an image of a facade and automatically receive accurate window measurements as output. These measurements can then be used to calculate the total wall surface.
Implementing this tool would drive innovation within the company, making it more technologically forward and attractive. It would also significantly reduce human error, streamline workflows, and boost overall efficiency by allowing employees to focus on more valuable tasks rather than repetitive manual labor.
Results
The project results have been structured in 5 different phases due to data quality issues. In the last phase it was decided to create a dataset with manual annotations based on the pictures Nelissen provided.
End Result:
Fully working webapp where a user can input a facade, it is segmentated by the model and posteriorly, the measurements of each window are returned based on user inputted scale.
Apart from this other deliverables include:
About the project group
Everyone in the group started with software profile and then assigned to AI-Core first and then AI-Advanced. Alpay did his internship in ASML. Aleksandar in ATHLON Netherlands. And Juan Alejandro in Legrand.
