Accurate sled velocity on a short-inclined track using accelerometer data

Wireless Inertial Measurement Units (WIMUs) are increasingly used to gather data and improve understanding of various human performance and complex motion scenarios. The Bob-Skeleton Push-Start features a stooped sprint from a crouch while pushing a heavy sled. Maximizing velocity during this brief period is considered crucial to performance, however it is poorly understood. An adjustable sled Push-Start training tool was instrumented with custom WIMUs, and a test subject performed 36 runs, with 12 combinations of 3 Incline and 4 Weight settings. A developed algorithm automatically identified, extracted, and integrated Pushing-Phase Acceleration data to Velocity and Displacement at hundreds of samples per second. Drift correction methods improved accuracy; while additional checks rejected problematic datafiles. WIMU derived Average Velocities were within - 0.005±0.074 meters per second (0.319±4.214%) of an existing Light-Gate system. Such an accurate, automatic, WIMU-based system could supplement or replace Light-Gate or other performance monitoring methods, while being more portable and readily usable by coaches or athletes. This would enable consistent, low-cost and high-fidelity, performance monitoring from the gym to the ice-track for improved candidate selection, comparison and training in Bob-Skeleton and other ice-track sled sports.