Study of road surface shape prediction for drivability evaluation of a mobility scooter in a grassy area

The aim of this study was to estimate the road surface shape in a grassy area in order to judge whether a mobility scooter can drive without accidents. For drivability evaluation, a road surface shape map is needed and it is constructed by a laser range finder (LRF). In order to make an accurate road surface shape map, it is necessary to obtain distance information of the road surface accurately. An experiment was conducted to understand characteristics when acquiring distance information in a grassy area. It was found that the distance information obtained from the LRF in a grassy area may include the information of the road surface. Therefore, we proposed a new road surface shape map expressed by the probability of being a road surface. An experiment was carried out to verify the validity of the proposed road surface shape map. It was found that the accuracy was sufficient for drivability evaluation.