@article{01K6ZDBRZJ67T3S2V5V4VFV3K8,
  abstract     = {{Orienteering exercises are an essential part of military training, especially in the context of infantry training. The completion of these exercises is often accompanied by poor performance and a large number of errors, for which supervisors often have no objective explanation. This makes evaluating these exercises very difficult even though this is an essential and non-trivial aspect within basic military training camps. Giving targeted feedback to each recruit is vital to their progress but still extremely difficult to facilitate. This article explores the possibility of assessing the performance of recruits during orienteering exercises by analyzing their chosen routes, providing insights which can lead to a better understanding of their success or failure. Exercises were recorded using a Garmin Fenix 7 and offloaded at scale through the Garmin Health SDK. We propose a methodology that leverages detailed terrain analysis and exercise-specific data to provide instructors with actionable insights into recruits' decision-making processes. Our technique that automatically detects checkpoints during the exercise achieved an F1-score of 94% and a mean displacement of 26.76 meters. This approach facilitates personalized feedback, enabling improvements in training outcomes. Additionally, the methodology can verify imposed constraints such as restricted areas or non-collaboration rules.}},
  author       = {{Decorte, Robbe and Criel, Dajo and Milleville, Kenzo and Slembrouck, Maarten and Verstockt, Steven}},
  issn         = {{1748-9725}},
  journal      = {{JOURNAL OF LOCATION BASED SERVICES}},
  keywords     = {{orienteering,wearables,military,map reading}},
  language     = {{eng}},
  number       = {{1-2}},
  pages        = {{99--111}},
  title        = {{Where is my recruit : assessing orienteering skills of military recruits through wearable data analytics}},
  url          = {{http://doi.org/10.1080/17489725.2025.2568373}},
  volume       = {{20}},
  year         = {{2026}},
}

Altmetric

View in Altmetric