INTRODUCTION: The choice of the racing line is a key success factor in alpine ski racing. Nevertheless, to define before the race a "proper" racing line for a given racing track remains a highly subjective process for coaches and athletes, as objective quality criteria for racing lines are not available. As such, the definition of racing lines and racing strategy is until now not accessible for scientific discussion opening up the room for myths and subjective rules-of thumb. The aim of this work is to establish a formal procedure for the computation of optimal racing lines before the race. With this, we hope to introduce objective factors supplementing the already available subjective factors. METHODS: On the basis of a physical model of an alpine ski racer and a 3D-measurement of the surface of the racing track, optimal control theory is applied to compute an optimal racing line for the ski racers. To this end, a nonlinear programming problem (NLP) is formulated with > 10`000 unknowns, and > 1 Mio. equality and inequality constraints, the solution of which is the optimal racing line. Additionally, the racing track is measured accurately with LIDAR- and photogrammetry- technology. With this, an accurate 3D-world could be created allowing to visualize the optimal racing line with Virtual-Reality-(VR)-technology, allowing for an highly immersive experience for the athletes. RESULTS/DISCUSSION: The AI-RACER has been tested on World-Cup racing tracks with professional athletes, after the competion has taken place. The error in forecasting the finish time was below 1.0 - 2.0 s. The accuracy of the physical model is surprisingly high for gliding phases. The model for more complex skiing techniques, such as drifting or energy pumping, needs further improvements, though. CONCLUSION: While not being perfect today, the AI-RACER opens up the space for advancements in the future. Clearly, there are promising future applications for racing as well as for training and education of young athletes.
© Copyright 2025 10th International Congress on Science and Skiing, January 28 - February 1, 2025, Val di Fiemme, Italy. Alle Rechte vorbehalten.