Master Thesis Proposal – AI & Vision for PCB Defect Detection

Do you want to explore the intersection of computer vision, AI, and edge computing?

In this thesis, we invite one or two students to develop methods for automatic defect detection on printed circuit boards (PCBs) using machine learning and exciting dataset. The project combines inference on both a resource constrained edge device and a powerful workstation, with traning performed on the workstation, enabling efficient deployment in industrial production environments.

About the project

The aim is to design and evaluate an AI-based vision system for PCB defect detection. Within this scope, you will:

• Work with existing PCB image datasets.
• Train, refine and benchmark different AI models (e.g. CNNs, transformer-based networks).
• Implement inference on edge device for real-time defect detection.
• Analyze system performance in terms of accuracy, latency, and resource efficiency.

Possible Research Questions

• Which AI-based vision architectures are most effective for detecting visual PCB defects?
• How can inference on edge hardware be optimized for visual PCB defect detection speed and accuracy?

Features & technologies

• Computer vision & deep learning (PyTorch, TensorFlow, OpenCV, Ultralytics)
• Edge inference
• GPU-accelerated training
• Industrial vision datasets

Who are you?

We are looking for one or two master’s students with an interest in AI, computer vision, and embedded systems. Experience with Python and machine learning frameworks is valuable, but the most important is that you are curious, motivated, and eager to explore AI in a real-world context. We also prefer you to have good knowledge in both Swedish and English.

To give you the best possible support during your thesis, we’d like you to be able to come to the office connected to the project and spend most of your time working from there.

Application:

We look forward to receiving your resume, and preferably, a personal letter in which you explain why you want to write your thesis with Syntronic.

We screen and evaluate applications on an ongoing basis. The thesis project may be filled before the application deadline.