Global navigation satellite systems such as the Global Positioning System (GPS) is one of the most important sensors for movement analysis. GPS is widely used to record the trajectories of vehicles, animals and human beings. However, all GPS movement data are affected by both measurement and interpolation errors. In this article we show that measurement error causes a systematic bias in distances recorded with a GPS; the distance between two points recorded with a GPS is – on average – bigger than the true distance between these points. This systematic ‘overestimation of distance’ becomes relevant if the influence of interpolation error can be neglected, which in practice is the case for movement sampled at high frequencies. We provide a mathematical explanation of this phenomenon and illustrate that it functionally depends on the autocorrelation of GPS measurement error (C). We argue that C can be interpreted as a quality measure for movement data recorded with a GPS. If there is a strong autocorrelation between any two consecutive position estimates, they have very similar error. This error cancels out when average speed, distance or direction is calculated along the trajectory. Based on our theoretical findings we introduce a novel approach to determine C in real-world GPS movement data sampled at high frequencies. We apply our approach to pedestrian trajectories and car trajectories. We found that the measurement error in the data was strongly spatially and temporally autocorrelated and give a quality estimate of the data. Most importantly, our findings are not limited to GPS alone. The systematic bias and its implications are bound to occur in any movement data collected with absolute positioning if interpolation error can be neglected.

In order to gain insight into skiing performance, it is necessary to determine the point where each turn begins. Recent developments in sensor technology have made it possible to develop simpler automatic turn detection methodologies, however they are not feasible for regular use. The aim of this study was to develop a sensor set up and an algorithm to precisely detect turns during alpine ski, which is feasible for a daily use. An IMU was attached to the posterior upper cuff of each ski boot. Turn movements were reproduced on a ski-ergometer at different turn durations and slopes. Algorithms were developed to analyze vertical, medio-lateral, anterior-posterior axes, and resultant accelerometer and gyroscope signals. Raw signals, and signals filtered with 3, 6, 9, and 12 Hz cut-offs were used to identify turn switch points. Video recordings were assessed to establish a reference turn-switch and precision (mean bias = 5.2, LoA = 51.4 ms). Precision was adjusted based on reference and the best signals were selected. The z-axis and resultant gyroscope signals, filtered at 3Hz are the most precise signals (0.056 and 0.063 s, respectively) to automatically detect turn switches during alpine skiing using this simple system.

No abstract

Introduction: Long-distance crosscountry skiing (XCS) has gained increased popularity within the past decades. However, research about long-distance XCS is limited; therefore, the aim of this study was to analyze the intensity distribution, technique application, and pacing strategies during long-distance XCS racing. Methods: Heart rate (HR) and section skiing speeds of 9 elite (ranked 1-100) and 10 amateur skiers (ranked 101-1,500) during the 90-km Vasaloppet race were collected. In addition, during the first uphill, the first 1,000 skiers were video-recorded to analyze the applied skiing strategy (e.g., grip-waxed skis versus exclusive double poling). Results: Mean race intensity was 82% of maximal HR and was not different between performance groups even though elite skiers skied ∼15% faster than amateurs. There was an interaction effect of section × group with a pronounced decrease in HR in amateurs compared with more even pacing in elite skiers (0.13 vs. 0.04% decrease/km) and skiing at higher percentage in the high-intensity zones in elite compared with amateurs (46 vs. 24%). Ninety-eight percent of the top 100 skiers and 59% of the first 1,000 skiers used exclusively double poling. Conclusion: Elite and amateur skiers ski at comparable mean race exercise intensity, but they have clear differences in skiing speed. The difference in the pacing profiles between elite and amateur skiers (more even vs. distinct positive pacing) demonstrate the greater capacity of the former with respect to physiological capacity and highlights that amateurs seem to start too fast according to their capacities. The exclusive application of the double poling technique is no longer a phenomenon of elite skiers but is widely used among the top 1,000 ranked skiers.

Volunteered geographic information (VGI) data-sets are characterised by heterogeneity due to influences from technical, social, environmental or economic factors. As a result, mapping progress does neither follow a spatially nor a temporally equal distribution, and thus can be hardly measured or predicted. Positively stated, heterogeneity leads to interesting VGI data-sets revealing regional peculiarities such as diverse community activities. This work proposes an approach for identifying regionally and temporally different developments with respect to mapping progress. Regional mapping progress is measured with a modified version of a previously proposed model for classifying activity stages, which has been used as foundation for a massive spatial and temporal analysis of the worldwide OpenStreetMap contributions between the years 2006 and 2013. It also allows the evaluation of rural and unpopulated areas. Results reveal that regional mapping progress heavily depends on a number of distinct influences such as geographical or legal borders, data imports, unexpected events or diverse community developments. The work highlights regions with distinct results by revealing individual mapping stories.

Several methodologies have been proposed to determine turn switches in alpine skiing. A recent study using inertial measurement units (IMU) was able to accurately detect turn switch points in controlled lab conditions. However, this method has yet to be validated during actual skiing in the field. The aim of this study was to further develop and validate this methodology to accurately detect turns in the field, where factors such as slope conditions, velocity, turn length, and turn style can influence the recorded data. A secondary aim was to identify runs. Different turn styles were performed (carving long, short, drifted, and snowplow turns) and the performance of the turn detection algorithm was assessed using the ratio, precision, and recall. Short carved turns showed values of 0.996 and 0.996, carving long 1.007 and 0.993, drifted 0.833 and 1.000 and snowplow 0.538 and 0.839 for ratio and precision, respectively. The results indicated that the improved system was valid and accurate for detecting runs and carved turns. However, for drifted turns, while all the turns detected were real, some real turns were missing. Further development needs to be done to include snowplow skiing.

In alpine skiing, four commonly used turning styles are snowplow, snowplow-steering, drifting and carving. They differ significantly in speed, directional control and difficulty to execute. While they are visually distinguishable, data-driven classification is underexplored. The aim of this work is to classify alpine skiing styles based on a global navigation satellite system (GNSS) and inertial measurement units (IMU). Data of 2000 turns of 20 advanced or expert skiers were collected with two IMU sensors on the upper cuff of each ski boot and a mobile phone with GNSS. After feature extraction and feature selection, turn style classification was applied separately for parallel (drifted or carved) and non-parallel (snowplow or snowplow-steering) turns. The most important features for style classification were identified via recursive feature elimination. Three different classification methods were then tested and compared: Decision trees, random forests and gradient boosted decision trees. Classification accuracies were lowest for the decision tree and similar for the random forests and gradient boosted classification trees, which both achieved accuracies of more than 93% in the parallel classification task and 88% in the non-parallel case. While the accuracy might be improved by considering slope and weather conditions, these first results suggest that IMU data can classify alpine skiing styles reasonably well.

No abstract