Detection of Lowering in Sport Climbing Using Orientation-Based Sensor-Enhanced Quickdraws: A Preliminary...

Area of Science:

• Sports Technology
• Sensor Systems
• Machine Learning

Background:

• Climbing gyms seek to enhance climber experience and optimize facility use.
• Tracking climbing sessions, particularly the descent, is key to analyzing performance and session completion.
• Existing methods may compromise climber privacy or be cost-prohibitive for gyms.

Purpose of the Study:

• To develop a privacy-preserving, cost-effective system for detecting climber descent in indoor climbing gyms.
• To analyze sensor data for patterns indicative of climber descent (lowering).
• To implement a machine learning model for automated descent identification.

Main Methods:

• A hardware prototype using energy-efficient accelerometer sensors attached to quickdraws was developed.
• Sensor data was collected in ultra-low power mode across various climbing routes.
• A supervised machine learning approach was employed, utilizing multidisciplinary feature engineering.

Main Results:

• Accelerometer sensors successfully captured distinct orientation patterns during climber descent.
• The developed supervised approach accurately identified lowering events.
• The system demonstrated energy efficiency, suitable for large-scale gym deployment.

Conclusions:

• The proposed sensor system offers a viable, privacy-conscious method for monitoring climber descent.
• This technology can provide valuable data for climbing gyms to understand climber behavior and improve services.
• Combining domain knowledge with machine learning enhances the effectiveness and practicality of sensor-based analysis in sports environments.