Supervisor: Univ.-Prof. Dr Elmar Rückert
Start date: 1st of July 2021
Theoretical difficulty: high
Practical difficulty: low
Sensor systems in cyber physical systems are highly complex with lots of different sensor types. Some of them work redundantly, utilizing different physical measuring effects. The goal of this thesis is to investigate sensor fusion by using spiking neural networks in movement planning. Therefore, the first step is to simulate two or more sensors with dynamical measurement noise in Matlab and subsequently try to find an optimal sensor fusion with the usage of spiking neural networks. The main objective is to draw the most out of the respective sensor types and to minimize the weaknesses as much as possible. This could be a key technology for autonomous robotics and driving.
 O. Bobrowski, R. Meir and Y. C. Eldar, “Bayesian Filtering in Spiking Neural Networks: Noise, Adaptation, and Multisensory Integration,” in Neural Computation, vol. 21, no. 5, pp. 1277-1320, May 2009, doi: 10.1162/neco.2008.01-08-692.
Implementation of spiking neural networks in robotics (sensory and planning):
 Xiuqing Wang, Zeng-Guang Hou, Feng Lv, Min Tan, Yongji Wang, “Mobile robots ׳†modular navigation controller using spiking neural networks”, Neurocomputing, Volume 134, 2014, Pages 230-238, ISSN 0925-2312, https://doi.org/10.1016/j.neucom.2013.07.055.
 Xiuqing Wang, Zeng-Guang Hou, Anmin Zou, Min Tan, Long Cheng, “A behavior controller based on spiking neural networks for mobile robots”, Neurocomputing, Volume 71, Issues 4–6, 2008, Pages 655-666, ISSN 0925-2312, https://doi.org/10.1016/j.neucom.2007.08.025.
 F. Alnajjar and Kazuyuki Murase, “Sensor-fusion in spiking neural network that generates autonomous behavior in real mobile robot,” 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), 2008, pp. 2200-2206, doi: 10.1109/IJCNN.2008.4634102.
Spiking neural networks and filtering:
 GILRA, Aditya; GERSTNER, Wulfram. Predicting non-linear dynamics by stable local learning in a recurrent spiking neural network. Elife, 2017, 6. Jg., S. e28295.