Absence from work caused by overloading the musculoskeletal system lowers the life quality of the worker and entails unnecessary costs for both the employer and the health system. Soft-robotic exoskeletons offer a possibility to overcome these problems by increasing the system flexibility, not limiting the supported Degrees of Freedom and being simultaneously an actuator and a joint. Since such exoskeletons can only be designed for using power when supporting the wearer, battery lifetime can be increased by covering only those actions for which support is needed. As regards controls, a major difficulty lies in finding a compromise between saving energy and supporting the wearer. However, an action-depending control can reduce the supported actions to only relevant ones and increase battery lifetime. The system conditions include detection of user actions in real time and distinguishing between actions requiring and not requiring support. We contributed an analysis and modification of human action recognition (HAR) benchmark algorithms from activities of daily living, transferred them onto industrial use cases and made the models compatible with embedded computers for real-time recognition on soft exoskeletons. We identified the most common challenges for inertial measurement unit based HAR and compared the best-performing algorithms using a newly recorded dataset of overhead car assembly for industrial relevance. By introducing orientation estimation, F1-scores could be increased by up to 0.04. With an overall F1-score without a Null class of up to 0.883, we were able to lay the foundation for using HAR for action dependent force support.
1. Hoy, D., Blyth, F. and Buchbinder, R. The epidemiology of low back pain. Best Pract. Res. Clin. Rheumatol., 2010, 24(6), 769–781.
https://doi.org/10.1016/j.berh.2010.10.002
2. Hagen, K. B. and Thune, O. Work incapacity from low back pain in the general population. Spine, 1998, 23(19), 2091–2095.
https://doi.org/10.1097/00007632-199810010-00010
3. Maniadakis, N. and Gray, A. The economic burden of back pain in the UK. Pain, 2000, 84(1), 95–103.
https://doi.org/10.1016/S0304-3959(99)00187-6
4. Gregorczyk, K. N., Hasselquist, L., Schiffman, J. M., Bensel, C., Obusek, J. P. and Gutekunst, D. J. Effects of a lower-body exoskeleton device on metabolic cost and gait biomechanics during load carriage. Ergonomics, 2010, 53(10), 1263–1275.
https://doi.org/10.1080/00140139.2010.512982
5. De Looze, M. P., Bosch, T., Krauze, F., Stadler, K. S. and O´Sullivan, L. Exoskeletons for industrial application and their potential effects on physical work load. Ergonomics, 2016, 59(5), 671–681.
https://doi.org/10.1080/00140139.2015.1081988
6. Gorissen, B., Reynaerts, D., Konishi, S., Yoshida, K., Kim, J.-W. and De Volder, M. Elastic inflatable actuators for soft robotic applications. Adv. Mater.,2017, 29(43), 1604977.
https://onlinelibrary.wiley.com/doi/abs/10.1002/ adma.201604977
https://doi.org/10.1002/adma.201604977
7. Polygerinos, P., Correll, N., Morin, S. A., Mosadegh, B., Onal, C. D., Petersen, K. et al. Soft Robotics: review of fluid-driven intrinsically soft devices; manufacturing, sensing, control, and applications in human-robot interaction. Adv. Eng. Mater., 2017, 19(12), 1700016.
https://doi.org/10.1002/adem.201700016
8. Ordóñez, F. J. and Roggen, D. Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition. Sensors, 2016, 16(1), 115.
https://doi.org/10.3390/s16010115
9. Plötz, T., Hammerla, N. Y. and Olivier, P. Feature learning for activity recognition in ubiquitous computing. In Proceedings of the 22nd International Joint Conference on Artificial Intelligence – Volume Two. IJCAI’11, Barcelona, Spain, July 16–22, 2011. AAAI Press, 1729–1734.
https://doi.org/10.5591/978-1-57735516-8/IJCAI11-290
10. Alsheikh, M. A., Selim, A., Niyato, D., Doyle, L., Lin, S. and Tan, H.-P. Deep activity recognition models with triaxial accelerometers. In CoRR abs/1511.04664, 2015.
http://arxiv.org/abs/1511.04664
11. Chen, Z., Zhang, L., Cao, Z. and Guo, J. Distilling the knowledge from handcrafted features for human activity recognition. IEEE Trans. Industr. Inform., 2018, 14(10), 4334–4342.
https://doi.org/10.1109/TII.2018.2789925
12. Kuschan, J., Goppold, J.-P., Schmidt, H. and Krüger, J. PowerGrasp: Concept for a novel soft-robotic arm support system. In Proceedings of ISR 2018; 50th International Symposium on Robotics, Munich, Germany, June 20–21, 2018. VDE, 269–274.
13. Shoaib, M., Bosch, S., Incel, O. D., Scholten, H. and Havinga, P. J. M. Fusion of smartphone motion sensors for physical activity recognition. Sensors, 2014, 14(6), 10146–10176.
https://doi.org/10.3390/s140610146
14. Banos, O., Villalonga, C., Garcia, R., Saez, A., Damas, M., Holgado-Terriza, J. A. et al. Design, implementation and validation of a novel open framework for agile development of mobile health applications. BioMedical Engineering OnLine, 2015, 14(26).
https://doi.org/10.1186/1475-925X-14-S2-S6
15. Leutheuser, H., Schuldhaus, D. and Eskofier, B. M. Hierarchical, multi-sensor based classification of daily life activities: comparison with state-of-the-art algorithms using a benchmark dataset. PloS One, 2013, 8(10), e75196.
https://doi.org/10.1371/journal.pone.0075196
16. Chavarriaga, R., Sagha, H., Calatroni, A., Digumarti, S. T., Tröster, G., del R. Millán, J. and Roggen, D. The opportunity challenge: a benchmark database for on-body sensor-based activity recognition. Pattern Recognit. Lett., 2013, 34(15), 2033–2042.
https://doi.org/10.1016/j.patrec.2012.12.014
17. Kunze, K., Barry, M., Heinz, E. A., Lukowicz, P., Majoe, D. and Gutknecht, J. Towards recognizing tai chi an initial experiment using wearable sensors. In Proceedings of the 2006 3rd International Forum on Applied Wearable Computing. VDE, 1–6.
18. Bao, L. and Intille, S. S. Activity recognition from user-annotated acceleration data. In Pervasive Computing (Ferscha, A. and Mattern, F., eds). Springer, Berlin, Heidelberg, 2004, 1–17.
https://doi.org/10.1007/978-3-540-24646-6_1
19. Blanke, U. and Schiele, B. Daily routine recognition through activity spotting. In Location and Context Awareness (Choudhury, T. et al., eds). Springer, Berlin, Heidelberg, 2009, 192–206.
https://doi.org/10.1007/978-3-642-01721-6_12
20. Rueda, F. M., Grzeszick, R., Fink, G. A., Feldhorst, S. and ten Hompel, M. Convolutional neural networks for human activity recognition using body-worn sensors. Inform., 2018, 5(2), 26.
https://doi.org/10.3390/informatics5020026
21. Murad, A. and Pyun, J-Y. Deep recurrent neural networks for human activity recognition. Sensors, 2017, 17(11), 2556.
https://doi.org/10.3390/s17112556
22. Skoda Dataset. Human Activity/Context Recognition Datasets.
http://www.har-dataset.org/doku.php?id=wiki:dataset (accessed 2019-07-26)
23. Walukiewicz, M. Aufbau und Inbetriebnahme eines Sensoranzugs für die messtechnische Erfassung von Posen und Bewegungen am Menschen. Bachelor thesis. Technical University of Berlin, Germany, 2019.
24. Kingma, D. P. and Ba, J. Adam: A method for stochastic optimization. 2014. arXiv:1412.6980 [cs.LG].
