The Embedded Sensing Group (ESG) addresses the crossroads of theory and practice in sensors and their communication-related aspects, exploring the different categories of sensors in a wide-range of real-world applications under the lens of performance, efficiency, and security. Also, we offer an up-to-date teaching curriculum on related aspects to both undergraduate and graduate students, partly in conjunction with other groups at the School of Computer Science (SCS) of the HSG.
Our group investigates how information flows from the physical world into the digital domain and the technology stack enabling this transition.
The research builds upon passive tracking and tracing systems, privacy-preserving perception methods, and secure cyber-physical infrastructures. Through practical deployments in diverse settings, including exhibition venues, healthcare facilities, and smart-home environments, we explored the effectiveness of passive sensing technologies, such as radio frequency identification, LiDAR, and multisensor fusion, in delivering actionable insights while safeguarding user privacy.
We also have an active collaboration with industry partners in the Vorarlberg region, and research groups within the SCS and beyond!
The ESG is committed to offer a full and up-to-date curriculum on embedded sensors and their communication aspects to both undergraduate and graduate students, partly in conjunction with other groups at the SCS.
We offer the following course(s) in Autumn:
Courses below are offered in the Spring:
Our group offers thesis projects at different levels, with scopes adapted to the student’s background and interests. We also welcome your topic proposals that align with our research areas. If you have a specific idea that aligns with our research, please get in contact (e-mail) to discuss supervision.
Develop a real-time system that listens for drones and estimates their 3D position using a network of microphones. You will record drone sounds, identify distinctive acoustic patterns, and apply signal processing to calculate positions under noisy conditions. The work includes hands-on flight tests and lays the groundwork for future counter-drone applications.
Design a web assistant that detects new IoT devices in a warehouse network, proposes safe segments using MUD profiles, and automatically applies VLAN and firewall rules. Using a Raspberry Pi for traffic monitoring and an AI-supported interface, you will validate that only necessary traffic is permitted while routine operations remain unaffected.
Investigate how combining acoustic and vibration sensors improves fault detection and precise localization in an operational pumped-storage hydropower plant. You will build a synchronized sensing pipeline, collect real-world data, and apply machine learning to identify the origin of anomalies in a live industrial setting.
Develop a gesture recognition system based on Time-of-Flight depth data that also verifies the identity of the authorized user from motion dynamics. You will implement and evaluate real-time performance in multi-user indoor scenarios without capturing personally identifiable visual details.
Explore how motion (PIR) and light (lux) traces in smart homes can reveal early signs of depressive states through routine and circadian changes. You will extend public datasets with clinically guided simulations, engineer behavioral features, and train lightweight models to distinguish depressive-like from normal periods.