IART: Inertial Assistant Referee and Trainer for Race Walking

Area of Science:

• Sports Science
• Biomechanics
• Wearable Technology

Background:

• Race walking requires precise technique to avoid infringements.
• Current methods for detecting race walking infringements and analyzing performance are limited.
• Objective analysis of biomechanical parameters is crucial for elite athlete development.

Purpose of the Study:

• To present the Inertial Assessment and Race-walking Technique (IART) system.
• To validate IART's capability in automatically detecting race walking infringements and analyzing sports performance.
• To assess the usability of IART for athletes, coaches, and referees.

Main Methods:

• Development of algorithms using raw inertial measurements from a single sensor at the L5-S1 vertebral column.
• Introduction of novel parameters: loss of ground contact time and step classification.
• Utilization of classic parameters: step length ratio, step cadence, and smoothness.
• Derivation of five biomechanical indices for elite athletes.
• Experimental protocol involving nine elite athletes over 300m paths.

Main Results:

• 99% average accuracy in detecting loss of ground contact events.
• 87% accuracy in classifying infringements per step sequence.
• 92% acceptable classification rate for infringements.
• Encouraging results regarding the accuracy of IART indices.
• Positive feedback on the user-centered design and intuitive radar chart representation.

Conclusions:

• IART demonstrates high accuracy in detecting race walking infringements and analyzing performance.
• The system offers practical usability for athletes, coaches, and referees.
• IART shows potential as a valuable tool for training enhancement and competition officiating.