Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Deep neural network based human activity recognition for the order picking process

: Grzezick, Rene; Lenk, Jan Marius; Rueda, Fernando Moya; Fink, Gernot A.; Feldhorst, Sascha; Ten Hompel, Michael


Yordanova, Kristina (Ed.); Schröder, M.; Bader, S.; Kirste, Thomas ; Association for Computing Machinery -ACM-; Association for Computing Machinery -ACM-, Special Interest Group on Computer and Human Interaction -SIGCHI-; Fraunhofer-Institut für Graphische Datenverarbeitung -IGD-, Darmstadt:
iWOAR 2017, 4th International Workshop on Sensor-based Activity Recognition and Interaction : 21. - 22. September 2017, Rostock
New York: ACM Press, 2017 (ACM International Conference Proceedings Series 1183)
ISBN: 978-1-4503-5223-9
Art. 14, 6 S.
International Workshop on Sensor-based Activity Recognition (iWOAR) <4, 2017, Rostock>
Fraunhofer IML ()
activity recognition; deep learning; convolutional neural network; inertial measurement unit

Although the fourth industrial revolution is already in progress and advances have been made in automating factories, completely automated facilities are still far in the future. Human work is still an important factor in many factories and warehouses, especially in the field of logistics. Manual processes are, therefore, often subject to optimization efforts. In order to aid these optimization efforts, methods like human activity recognition (HAR) became of increasing interest in industrial settings. In this work a novel deep neural network architecture for HAR is introduced. A convolutional neural network (CNN), which employs temporal convolutions, is applied to the sequential data of multiple intertial measurement units (IMUs). The network is designed to separately handle different sensor values and IMUs, joining the information step-by-step within the architecture. An evaluation is performed using data from the order picking process recorded in two different warehouses. The influence of different design choices in the network architecture, as well as pre- and post-processing, will be evaluated. Crucial steps for learning a good classification network for the task of HARin a complex industrial setting will be shown. Ultimately, it can be shown that traditional approaches based on statistical features as well as recent CNN architectures are outperformed.