Jona Eisele, M.Eng.

Industrial PhD

Ultrasonic environmental sensing based on deep learning methods

There is a growing need in advanced environmental sensing for autonomous driving applications. Improving the performance of ultrasonic sensors is of great interest because of their robustness, low production costs and widespread use. Today’s sensors use the pulse-echo method to calculate the distance to obstacles. This work examines methods of signal processing and feature extraction as well as the use of machine learning to perform ultrasound-based classification of objects.

Research Topics

  • Ultrasound
  • Machine Learning
  • Deep Learning

Project Partners

  • Technical University of Munich
  • Stuttgart Media University
  • Robert Bosch GmbH



  • Eisele, Jona; Gerlach, André; Maeder, Marcus; Koch, Andreas; Marburg, Steffen: Objektklassifikation auf Basis von Luftultraschall für Aufgaben der Umfeldsensierung mittels Convolutional Neural Network. Fortschritte der Akustik - DAGA 2023, 49. Jahrestagung für Akustik, 06.-09. März 2023, Hamburg, Deutsche Gesellschaft für Akustik e.V. (DEGA), 2023 more… BibTeX
  • Eisele, Jona; Gerlach, André; Maeder, Marcus; Marburg, Steffen: Convolutional neural network with data augmentation for object classification in automotive ultrasonic sensing. The Journal of the Acoustical Society of America 153 (4), 2023, 2447-2459 more… BibTeX Full text ( DOI )


  • Eisele, Jona; Gerlach, André: Künstliche Intelligenz für akustische Sensorsysteme. Fortschritte der Akustik - DAGA 2022, 48. Jahrestagung für Akustik, 21.-24. März 2022, Stuttgart, Deutsche Gesellschaft für Akustik e.V. (DEGA), 2022 more… BibTeX