Skilog: A Smart Sensor System for Performance Analysis and Biofeedback in Ski Jumping
Skilog wearable biofeedback system
Abstract
Skilog is a wearable system designed to assist ski jumpers by analyzing take-off mechanics and providing real-time biofeedback. Pressure-sensitive insoles in each boot sample at 100 Hz and stream to an embedded GAP9 microcontroller, where an XGBoost model classifies the athlete’s jump phases. The system achieves 92.7% accuracy in classifying jumps and offers immediate feedback on timing and technique while weighing about 20 g and consuming little power, enabling multi-day operation.
Key Highlights
- Pressure-sensitive insoles feeding an XGBoost classifier deliver 92.7% jump-phase accuracy for ski jumping.
- Runs entirely on a GAP9 MCU with low power consumption, enabling multi-day operation from a lightweight hardware setup.
- Provides immediate biofeedback to athletes and coaches to refine take-off timing and technique.
Thorir Mar Ingolfsson
Postdoctoral Researcher
I develop efficient machine learning systems for biomedical wearables that operate under extreme resource constraints. My work bridges foundation models, neural architecture design, and edge deployment to enable real-time biosignal analysis on microwatt-scale devices.