Author(s): THORWARTL, C., VON SCHLEINITZ, J., STOECKL, F., KREMSER, W., STAUDINGER, R., GRAH, T., SANDER, A., Institution: SALZBURG RESEARCH, Country: AUSTRIA, Abstract-ID: 1454

INTRODUCTION:
Recent work introduced the concept of a "smart luge", demonstrating that steering actions can be captured using force sensors placed at key interaction points between athlete and sled [1]. While these studies focused on feasibility, quantitative models linking athlete input to runner contact behavior are still lacking in the literature. Such models are relevant, as runner deformation and contact conditions directly influence ice friction and sled dynamics.
The aim of this pilot study was to develop and apply a fully instrumented luge to capture athlete force input, runner deformation, inter-runner spacing, and runner-ground contact mechanics, and to establish a quantitative input-output model of sled steering. We hypothesized that shoulder and bow forces would systematically predict runner contact center position and contact length.
METHODS:
A luge was instrumented with force sensors at the shoulders and bows, strain gauges on both runners (front section) to capture runner deformation, and linear potentiometers to quantify the relative distance between the left and right runners. Runner-ground contact mechanics were derived from stabilized video recordings based on the absence of transmitted light underneath the runner, yielding contact length and contact center position for each runner.
In a controlled laboratory setup, an experienced former Olympic-level athlete performed 20 quasi-static steering cycles at three steering intensities in simulated left and right turns. Signals were segmented into steering cycles and time-normalized. Standardized multiple regression analyses using athlete force inputs were applied to quantify input-output relationships between athlete forces and mechanical contact measures.
RESULTS:
Contact center position was explained almost completely by athlete force inputs alone, with adjusted R2 values ranging from 0.93 to 0.98. Standardized regression coefficients showed direction-dependent dominance, with bow forces prevailing in left turns and shoulder forces in right turns (absolute beta = 0.80-0.94, all p < 0.001). In contrast, contact length showed lower and more variable explainability and did not exhibit systematic modulation across steering conditions.
CONCLUSION:
This study demonstrates the feasibility of a fully instrumented luge for quantitative input-output modeling of sled steering. Steering-related contact mechanics, particularly contact-center modulation, can be described by a compact model driven primarily by athlete force application. The observation that direction-dependent force dominance occurred even in an experienced former Olympic-level athlete suggests that distinct steering strategies may be applied implicitly rather than consciously and could be addressed in simulator-based and dry-land training using objective force-based feedback.
References
1. Staudinger R, Kremser W, Thorwartl C, Schwameder H. Towards a smart luge that measures steering input of the rider. Proc 12th Int Conf Sport Science Research and Technology Support. 2024.