As Austrias largest research and technology organisation for applied research we are dedicated to make substantial contributions to solving the major challenges of our time climate change and digitalisation. To achieve our goals we rely on our specific research development and technology competencies which are the basis of our commitment to excellence in all areas. With our open culture of innovation and our motivated international teams we are working to position AIT as Austrias leading research institution at the highest international level and to make a positive contribution to the economy and society.

Our Center for Vision Automation & Control (VAC) located in Vienna invites applications for a masters thesis position. VAC exploits the opportunities of automation and digitalisation to develop intelligent solutions in computer vision robotics and control. We work closely with industrial and academic partners to bring cutting-edge algorithms onto real machines and into real environments.

As part of the Competence Unit Assistive & Autonomous Systems (AAS) we research safe and dependable autonomous systems and mobile work machines for challenging outdoor and industrial scenarios. We focus on robust perception localisation planning and monitoring technologies that enable machines to operate reliably reduce workload for human operators and increase overall system safety.

• Dive into the state of the art in vibration- and acoustic-based condition monitoring anomaly and event detection and identify promising methods.
• Prototype on established open datasets implementing and benchmarking AI-based anomaly and event detection methods (e.g. feature-based models one-class approaches autoencoders lightweight CNNs) using vibration and/or machine-sound data.
• Design a compact sensor concept for a real work machine selecting suitable vibration and/or audio sensors and defining mounting positions on or near the critical mechanical structures.
• Support data acquisition on a test platform or test rig recording and labelling normal operation under different loads and operating modes as well as selected anomaly scenarios (e.g. imbalance loosened non-critical components controlled impact / foreign-object events).

• Develop end-to-end detection pipelines from signal preprocessing (filtering feature extraction time-frequency representations) to model training and thresholding for online anomaly and event detection.
• Systematically evaluate and compare your models analysing detection rates false alarms and robustness to operating conditions and noise and assessing the impact of different sensor configurations (vibration vs. audio vs. sensor fusion) and model choices.
• Work as part of an experienced research team receive close mentorship from scientists in robotics machine learning and control and learn how to plan document and present an applied research project in a professional R&D environment giving your scientific career a strong application-driven kick-start.