As part of the “Lev4ISM” research project, work is being carried out on an ultrasonic levitator system that is intended to enable non-contact handling of matter. Samples of different sizes must be moved simultaneously and precisely in the levitator’s acoustic field. Furthermore, the control should allow high actuation speeds while preventing the sample from falling out of its acoustic trap. A stereo camera system is therefore used as feedback, which tracks the samples and returns the position in real time.

• Literature review on suitable controller structures
• Development of a concept for real-time control of transducer arrays
• Development of a controller & implementation in existing software
• Validation of the controller through experiments

• Basic programming experience
• Knowledge of controller design desirable, but not mandatory

As part of the Lev4ISM research project, work is being carried out on an ultrasonic levitation system that will enable contactless handling of matter in different gravitational environments. This includes liquids, which must be automatically introduced into the levitator's sound field. To this end, an acoustic system is to be developed that uses sound pressure fluctuations to extract nanoliter-sized droplets from a liquid reservoir. These droplets are then to be collected in a second acoustic field.

• Literature research on suitable methods of fluid excitation
• Construction of a suitable transducer array
• Development of an algorithm for controlling the transducers
• Implementation of the algorithm in an existing control system
• Implementation of a method for collecting the ejected droplets
• Validation of the process through experiments

• Basic programming experience
• Working independently

As part of the "Lev4ISM" research project, work is being carried out on an ultrasonic levitator system that will enable the contactless handling of matter. Samples of different sizes must be moved simultaneously and precisely in the levitator's sound field. This requires control at high actuation speeds. As feedback for the controller, a tracking system is to be developed as part of this thesis, which tracks the samples on the basis of a stereo camera system and returns the position in real time.

• Literature research on suitable tracking algorithms and hardware
• Potential analysis for the use of machine learning to minimize latencies in tracking
• Visualization of the tracked objects in a virtual environment
• Optimization of the currently used image processing algorithms and their extension by simultaneous multi-object recognition
• Experimental validation of the real-time capability and accuracy of the tracking system

• Basic programming experience in C++
• Knowledge of passive marker tracking and its visualization desirable, but not mandatory