A Wearable Assistant for Snowboard Training


Learning new sports is often difficult and time consuming. Students need to practice for a long time until they correctly perform the techniques of a sports domain. In snowboarding, instructors cannot be with students while going down the slope. Due to this spatial separation, a trainer typically provides feedback after exercises. Our wearable system is intended to automatically supervise posture during the ride and to alert users to incorrect body movements. Tiny sensors on the rider's body detect common snowboarding mistakes. Actuators provide immediate tactile motion instructions as feedback that subtly communicates hints for corrections.


We developed a wireless prototype system that senses the rider’s motion and posture on the snowboard. The system consists of a custom-built sensor & actuator box (hosting one Bluetooth Arduino board), two Bluetooth Shake SK6 inertial sensor packs, two bend sensors, and four force-sensitive resistors (FSR). We used a Nokia N70 mobile phone as host device running a Python script to sample sensor data over Bluetooth. The bend sensors were attached to the back of each knee to detected knee flexion of the rider. To measure weight distribution on the snowboard two FSRs were inserted into each boot. The digital compass algorithm of the Shake devices was used to measure the rotation of the rider’s upper body relative to the rotation of the snowboard.

User Study

Interviews with snowboard instructors confirmed that a wearable system that automatically senses the rider's posture, detects riding mistakes, and provides real-time feedback during descents, might be very useful to teach snowboarding. A pilot study in an indoor winter sport resort showed that it is possible to accurately detect insufficient knee bending and to estimate the weight distribution on the snowboard in real-time.


Created by spelmezan. Last Modification: Monday 21 of September, 2009 12:55:29 by dennis.

Media Computing Group at RWTH Aachen