Title: Situational Awareness for Tactical Applications
Author(s): L. Ruotsalainen, R. Guinness, S. Gröhn, L. Chen, M. Kirkko-Jaakkola, H. Kuusniemi
Published in: Proceedings of the 29th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2016)
September 12 - 16, 2016
Oregon Convention Center
Portland, Oregon
Pages: 1190 - 1198
Cite this article: Ruotsalainen, L., Guinness, R., Gröhn, S., Chen, L., Kirkko-Jaakkola, M., Kuusniemi, H., "Situational Awareness for Tactical Applications," Proceedings of the 29th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2016), Portland, Oregon, September 2016, pp. 1190-1198.
Full Paper: ION Members/Non-Members: 1 Download Credit
Sign In
Abstract: This paper presents the results of detecting various motion states, essential for infrastructure-free tactical situational awareness, using Machine Learning (ML) for motion classification. We investigated if the use of multiple IMUs will bring additional information and improve the sensing of motion contexts. Namely, we used three different Inertial Measurement Units (IMUs) with the first attached to the user’s torso, the second to the foot, and the third to the helmet. We also studied the best combination of sensors, measurements, and features used for detecting the contexts and for obtaining an accurate position solution. The full set of sensors studied included the three IMUs, a camera, a barometer, and an ultrasonic sensor. The method was tested with data collected in two experiments encompassing various motion patterns. Results showed that all sensors bring added value to the motion detection and that the motions having multiple instances in the data, even the ones considered as difficult to be identified like crawling, could be accurately classified. Also, the selection of features used for the classification process is discussed and evaluated as well as few different ML algorithms for classification. This research is a significant step towards infrastructure-free situational awareness for tactical applications.