DC03 Jasmien Hassanien- IZFP
Autonomous sensors are a promising technology for monitoring of critical infrastructure such as bridges, pipelines, etc, but also autonomous production processes. However, widespread adoption of the concept is still impeded by major practical hurdles, such as the availability of suitable low-power electronics that allow off-grid operation over extended periods of time as well as on-chip data reduction techniques that reduce the communication load for remote-location operation. Generic data reduction approaches like Compressed Sensing (CS) might provide a suitable approach to deliver reusable architectures that can be manufactured at the required scales. Compressed Sensing for mechanical wave sensors is still in a conceptual phase (TRL 2). The solution is based on the fact that prior knowledge on the signals is often available in monitoring applications, due to the repetitive nature of the measurement.
Keywords: Compressed Sensing, sub-Nyquist sampling, infrastructure monitoring, active and passive sensing, ultrasonic imaging.
Research fields: Electrical engineering
Click here to watch Jasmien introducing herself as well as her PhD subject.
The objective of the project is to explore solutions to data reduction techniques for monitoring applications in the scope of the applications focused by partners of the USES² consortium, i.e., with a focus on acoustic/ultrasonic sensors. Through tight interaction with the corresponding partners, a thorough comprehension of the relevant sensor modalities should be established so that a clear understanding of relevant and irrelevant parts of the measurement signals is achieved. Based on this, concepts and architectures for data reduction should be researched. This process follows three basic steps: (a) software-based data reduction (post acquisition), (b) on board data reduction close to the sensor (e.g. FPGA-based), (c) on-chip data reduction (via suitable custom electronic design). The concept will be evaluated using concrete applications within the USES² consortium.
1. Simulation framework and procedure for the numerical validation of compression techniques applied to sensor data
2. Investigations at the level of applicability of compression with regard to the availability of prior knowledge and its quality
3. Proof of concept in the form of a lab demonstrator
4. Conclusion on the industrial applicability of the CS approach with respect to real-world performance metrics (robustness, power budget, electronic design cost) and constraints (security, latency, etc.).
- UEiffel, Odile Abraham, 2 months, M13-M14: Joint work in (sub-Nyquist) measurement design for NCWI.
- Airbus, Jaime García-Alonso, 2 months, M20-M21: Joint work on the effect of EOC variations on sub-sampling for SHM.
- UPM, Antonio Fernandez-Lopez, 2 months, M29-M30: Investigation of CS concepts for optical fibre sensing interrogation.
- J. Kirchhof, S. Semper, C. Wagner, E. Pérez, F. Römer, and G. Del Galdo, Frequency Sub-Sampling of Ultrasound Non-Destructive Measurements: Acquisition, Reconstruction and Performance, in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 68, issue 10, pp. 3174 – 3191, Oct 2021, doi:10.1109/TUFFC.2021.3085007, preprint: https://arxiv.org/abs/2012.04534
- O’Connor, S. M., Jerome P. Lynch, and A. C. Gilbert. "Compressed sensing embedded in an operational wireless sensor network to achieve energy efficiency in long-term monitoring applications." Smart Materials and Structures 23, no. 8 (2014): 085014, https://iopscience.iop.org/article/10.1088/0964-1726/23/8/085014/pdf
- Zonzini, Federica, Antonio Carbone, Francesca Romano, Matteo Zauli, and Luca De Marchi. "Machine learning meets compressed sensing in vibration-based monitoring." Sensors 22, no. 6 (2022): 2229, https://mdpi-res.com/d_attachment/sensors/sensors-22-02229/article_depl…
Contact
Contact
DC03 Jasmien Hassanien
Contact
Local academic supervisor Florian Roemer (IZFP)
Contact
Industrial co-supervisor Jaime Garcia Alonso (Airbus)
Contact
PhD supervisor Giovanni Deelgaldo (TUIL)