Characterization and Classification of Tactical Movements Using Wearable Motion Sensors and Deep Learning Models

This work presents the first comprehensive IMU-based human activity characterization and classification of 16 operationally relevant tactical movements, such as crawling, jumping through a window, and climbing over a wall. Four active-duty soldiers completed 16 tactical tasks while wearing full-body inertial measurement units (IMUs), and lower-limb joint kinematics were analyzed to inform design requirements for wearable assistive devices for tactical operations. Two deep learning models, including an attention-based Transformer with self-supervised pretraining, were implemented and achieved up to 95% accuracy. Additionally, six sensor configurations ranging from one to three IMUs were evaluated to determine the minimum number of wearable motion sensors and their optimal placement strategies. Cross-validation and leave-one-subject-out testing confirmed the models’ robustness and generalization. These findings demonstrate the feasibility of classifying tactical movements in real-world environments and offer guidance for sensor placement and assistive device design.