Abstract
Background
The performance standard in snowboard slopestyle competitions is very high, as shown by their inclusion in the Winter Olympic program. Snowboard slopestyle is acrobatic in nature, with athletes combining technical tricks and skilful riding. Tricks involve athletes spinning in all three rotational axes (longitudinal, frontal and sagittal) while riding down a purpose-built course. There is a paucity of research into the training of snowboard slopestyle athletes. Common training descriptors such as volume, intensity and overall training load cannot be used for on-snow training, as there is currently no comprehensive athlete tracking system. A literature review highlighted the need for research into the development of a snow sport-specific athlete tracking system, potentially using wearable inertial measuring units (IMUs). Such emerging technologies could empower coaches, athletes, strength and conditioning practitioners and physiotherapists, giving them a greater insight into the physical demands of training and competition.
Aims
The overall aim of this thesis was to develop an athlete tracking tool that could be used in the daily training environment of elite snow sports athletes. Current technologies used in athlete tracking were explored and their potential application to track the movements of snowboard slopestyle athletes on-snow were identified. Data analysis techniques were developed to enable quantification of key training variables and allow coaching staff to monitor athlete training load. Such tracking could improve athletes’ performance in competition by helping to accelerate progression, limiting the negative effects of over- or under-training and time lost due to injuries. To address these aims, the thesis was constructed around a number of independent but linked research projects.
Methods
A sequential multi-study approach was used to identify the technologies and data analysis techniques required for the development of an automated athlete tracking tool. A critical literature review identified current methods of athlete tracking and how they are used within a wider athlete monitoring program. IMUs containing an accelerometer, gyroscope, magnetometer and global positioning system sensors were recommended for tracking athletes on-snow. Validity and reliability testing identified the measurement error of the gyroscope and accelerometer sensors in each IMU. A method for attaching the IMUs to the athletes was developed and used for laboratory-based testing of landing impacts and trick identification. Data analysis techniques identified during the thesis were used to develop an automated athlete tracking tool. Initial field testing gave insight into the application and feasibility of tracking athletes on-snow. The efficiency and accuracy of the athlete tracking tool was then examined. Four elite snowboard slopestyle athletes participated in a two-week block of training, wearing an IMU that allowed for post-session analysis. Information gathered from the on snow validation process was used to refine data analysis techniques and inform the development of the athlete tracking tool. Over a three-month period, four elite New Zealand snowboard slopestyle athletes that are part of Snow Sports New Zealand (SSNZ) High Performance Programme, were tracked during training and competition. Automated on-snow training reports were generated, identifying key training variables of time on course; count of runs; count of jumps and rails, classification of landing impacts; and tricks performed. Analysis of the training reports along with collected session rate of perceived exertion scores was used to describe different phases of an athlete’s training plan.
Results
During the 3-month tracking of snowboard slopestyle athletes, a total of 36 days of training were tracked, within which 561 runs were completed with a mean of 14.6 runs per day. A total of 4015 air time events were detected, which were split into jumps (2641) and rails (1374). Mean jumps and rails per day were 41.8 and 38.2 respectively. The longest recorded time on-snow was 4hrs and 58mins and the shortest time was 27 minutes, air time over jumps and rails was similar (1.5s and 1.6s respectively); however, larger angular displacements were observed on jumps. Session rate of perceived excursion (RPE) scores varied over the course of the training season with a mean of 5.9 (minimum of 1 and maximum of 9) for the athlete group. Aerial tricks that involved angular displacement (rotations) ranged from 180° to 1620°. Straight airs (less than 90° of angular displacement) were completed most frequently, with 967 straight airs over jumps and 746 over rails. Rails showed a smaller range of angular displacement than jumps, with tricks ranging from 180° up to 720°. Spin direction and rider orientation at take-off indicated that forward spins in both frontside and backside directions were completed most frequently (30% and 33% respectively). Landing analysis from on-snow accelerometer data, indicated that 4013 landings took place, with an average of 75 landings per training session. Landing impulse measures observed a mean of 385 g·s with 95% CI [392, 375]. The largest landing impulse recorded was 1387 g·s. Landing time was normally distributed, with 33.0% of landings occurring between 0.06 s and 008 s. Of the total landings, 774 were classified as heavy or very heavy landings.
Conclusion
The athlete tracking tool developed in this study showed which training variables could be extracted to monitor training load. Differences between competition and training were observed; competitions had fewer runs but heavier landings, while during training, athletes were on-snow for longer and completed more jumps and rails. During a high-intensity training camp fewer heavy landings were observed when the athlete was landing on a landing bag and more difficult tricks were performed with increased frequency when compared to regular training. Monitoring training load while returning an athlete from injury is vital for reducing the risk of re-injury. Information gathered from the athlete tracking tool, was used to inform return to snow policies for the wider SSNZ program.
Having the ability to monitor snowboard slopestyle athletes throughout a season added much-needed on-snow load monitoring data to already established off-snow monitoring tools used in the high-performance programme. It is recommended that along with the objective quantification of training variables, RPE scores be included in the daily monitoring of snowboard slopestyle athletes. Future use in a programme relies on the timely feedback of information. Therefore, the athlete tracking methods, data analysis and feedback techniques developed within the current research project should be integrated into the SSNZ high performance program and form part of the athlete monitoring program. Future development of the athlete tracking tool would enable other SSNZ athletes to benefit from on-snow monitoring; modifications to the data analysis tool would enable Freeski slopestyle and halfpipe athletes to be monitored on-snow. As a high-performance program all park and pipe athletes would be able to benefit from this research where athletes’ progression, fatigue and work load can be managed to improve performance and reduce the risk of injury.