Thesis project: Defect detection for foundries using advanced computer vision systems

Introduction:
Casting defects are an inevitable part of the casting process where a liquid material is poured into a mould that contains a hollow cavity of the desired shape. There are many kinds of casting defects, including blow holes, pinholes, burrs, shrinkage flaws, flaws in the mold material, pouring flaws, metallurgical flaws, etc. One the key issues has been the time-consuming inspection process which is often carried out manually often leading incorrect rejection due to human accuracy. By utilizing automated quality inspection systems, companies can reduce the time taken to identify and reject defective products, maximizing thus the accuracy of the process.

Computer vision is an area of research that has seen tremendous growth in recent years, thanks to advances in artificial intelligence (AI). In the proposed thesis, the student will be tasked with training and evaluating an industrial-grade AI CV system for surface defect detection from a well-known commercial vendor, at the cast iron foundry at Baettr Guldsmedshyttan. The system has several in-built tools for image annotation, and includes a library of AI models, but model selection and training must be done on-site by the end user. Several approaches will be evaluated, as different categories of AI CV models, and different ways of treating the training data, may have distinct advantages and disadvantages with respect to defect detection.

Research goals:
The goal of this thesis project is to improve the performance of a commercial, AI-based computer vision system that can already accurately recognize many defects that are typically found on the surfaces of cast metal components. Tasks include

• identification of key defects and dataset collection
• investigating the need for and potentially generating synthetic datasets
• training and evaluating suitable AI CV models, such as object detection and image segmentation models
• deploying the models and evaluating their performance in real production

Methodology:
The proposed research project will use a CV-system. Herein, various machine learning (ML) techniques will be evaluated for improvement of the CV-system. Additionally, the CV-system will be trained on different types of data focusing on the malformations, cracks and/or discolorations.

General information about the project:

Start and end date: VT 2025 (january-june)

Degree level: The project is preferably carried out at master’s level or similar but can also be suitable at bachelor’s level.
Place: partly flexible. Practical parts of the work must be done at the foundry in Guldsmedshyttan.

Supervisors at Baettr:
Jörgen Säldefjord, Baettr, [email protected]

Supervisors at RISE
Andreas Thore, RISE, [email protected]
Lennart Elmquist, RISE, [email protected]

About Baettr
We work every day to make wind energy the most widely available and preferred source of energy in the world. Because we believe that passing on a better place for future generations is not only profitable – it is truly enriching. We cooperate with international OEM’s operating in the wind industry, within our full spectrum of services: Design, casting, machining, surface treatment and assembly.

Now you have the opportunity to join us at the foundry in Guldsmedshyttan within the SMYG-project. The foundry was originally founded in the 15th century, so to this day, Guldsmedshyttan serves as one of the oldest operating foundries in the world. Today we’re approximately 150 employees working at facility and working towards our moto: Engineering the foundation for future generations

When you are joining us for our thesis you have the possibility to lend an apartment close by the foundry and you will also get a financial compensation for your work.

Place of work:
Elzwiks väg 1
71178 Guldsmedshyttan

Application deadline:
4. January 2025

Apply

Basic Information