Multimodal Sensing the Human-in-the-Loop Haptic Dynamics of the Trombone Handslide

The trombone is both an iconic musical instrument and a human-in-the-loop haptic feedback device. During performance, the trombone handslide provides multimodal feedback to the musician, enabling precise motion control. Yet the haptic dynamics of the slide have scarcely been studied beyond a kinematic perspective. This work introduces a multimodal sensor test rig that measures translational forces and slide motion in real-time, integrating a load cell and accelerometer at the musician's grip with an ultrasonic motion sensor. Experiments were performed across several musical scales, and the test rig's performance was evaluated in terms of sensor noise, latency, and positional drift. One-dimensional time-series data reproduced expected stair-step position profiles, while higher-dimensional dynamics revealed semi-elliptical trajectories that reflect symmetry between compressive and tensile applied forces. Consistent with the nonlinear dynamics observed throughout the force–motion data, preliminary system identification indicated that linear models cannot accurately predict position from force. Based on these findings, a haptic control policy was proposed within an optimal feedback control framework, wherein the trombonist regulates the applied force so that the sensed trajectory tracks a learned reference while minimizing tracking error and actuator effort. Limitations of the test rig include positional drift from secondary body motion, millimeter-scale spatial resolution, and limited range of motion due to the hardware configuration. Despite these constraints, the multimodal test rig and dataset provide a foundational starting point for future work, including nonlinear system identification, dynamic friction modeling, and exploration of human-in-the-loop control and learning mechanisms.