Event Camera-Based In-Situ Quality Monitoring for Additive Manufacturing (LPBF) using Deep Learning

This topic presents an internship and thesis opportunity focused on developing an event camera-based in-situ quality monitoring system for the Laser Powder Bed Fusion (LPBF) additive manufacturing process using deep learning techniques.

• Event camera advantages for LPBF: Event cameras provide asynchronous output only on intensity changes, enabling microsecond latency, high temporal resolution, and resistance to overexposure, which are beneficial for capturing the fast and bright melt pool dynamics in LPBF manufacturing.

• Project goals and methodology: The project aims to design and validate a quality monitoring framework by collecting event camera data under varied LPBF conditions, creating novel event-based data representations, and training deep learning models such as CNNs or RNNs for real-time defect classification.

• Learning outcomes: Participants will gain skills in using event camera systems in challenging industrial environments, designing deep learning architectures for sparse asynchronous data, applying machine learning to industrial quality monitoring, and performing experimental validation on real LPBF testbeds.

Suitable candidates hold a bachelor's degree in engineering or computer science with knowledge of deep learning and image processing, are proactive, communicative in English, and passionate about research.