Deep Learning-Based Travel Time Estimation in Hiking with Consideration of Individual Walking Ability

Abstract Hiking is popular, but mountain accidents are serious problems. Accurately predicting hiking travel time is an essential factor in preventing mountain accidents. However, it is challenging to accurately reflect individual hiking ability and the effects of fatigue in travel time estimation. Therefore, this study proposes a deep learning model, “HikingTTE”, for estimating arrival times when hiking. HikingTTE estimates hiking travel time by considering complex factors such as individual hiking ability, changes in walking pace, terrain, and elevation. The proposed model achieved significantly higher accuracy than existing hiking travel time estimation methods based on the relation between slope and speed. Furthermore, HikingTTE demonstrated higher accuracy in predicting hiking arrival times than a deep learning model originally developed to estimate taxi arrival times. The source code of HikingTTE is available on github for future development of the travel time estimation task.