Open Position @ GKmM

Research Topic: Composite Event Detection and Processing in Human-Worn and Wireless Sensor Networks Surveillance scenarios frequently contain a highly variable and heterogeneous set of sensor modules such as camera footage, passive infrared information, beam break sensors, and positioning technology, that all report their data intermittently to human-operated clusters. The topic of this PhD thesis adds, apart from a wider range of sensor types and number, the inclusion of personal sensor nodes acting within body sensor networks of humans active in the environment (such as rescue workers in a search and rescue scenario). Composite event detection builds upon other work realized in the Research Training Group, to detect complex events such as hazardous situations that cannot be measured on individual nodes only, by combination of locally detected events (e.g., detections of spreading gas and fire leaks in each other’s proximity).

This PhD work will involve the study of efficient pattern recognition and motif discovery algorithms to detect events in time-critical scenarios, assuming a network of well-located computing nodes with limited resources. A healthy amount of the work will be dedicated to evaluating the achieved work in realistic and stressful environments. The student will work together with several experienced PhD students in the same area and should be interested in non-trivial signal processing and machine learning, as well as embedded systems. The topic will be supervised by Kristof Van Laerhoven in the ESS group at the Technische Universität Darmstadt.