Summary
International Conference on Emerging Technologies for Communications
2022
Session Number:O3
Session:
Number:O3-1
Spatio-temporal model that aggregates information from sensors to estimate and predict states of obstacles for control of moving robots
Yuichi Ohsita, Shinya Yasuda, Taichi Kumagai, Hiroshi Yoshida, Dai Kanetomo, Masayuki Murata,
pp.-
Publication Date:2022/11/29
Online ISSN:2188-5079
Summary:
Mobile robots need to be controlled to perform the required tasks without collisions with obstacles. However, there are moving obstacles such as other robots and people in the work area of the mobile robot. To control the robot efficiently without collisions with such obstacles, it is important to predict the risk of collisions at each location and each time. In this paper, we propose a framework that aggregates the information from sensors and predicts future states at each location to calculate the risks. This framework is based on the spatio-temporal model. In this framework, we introduce a manager that maintains the spatio-temporal model. We also deploy sensors in the environments. Then sensors send the monitored information to the manager. The manager maps the monitored information to the spatio-temporal model and updates the current and predicted future states. In this paper, we also demonstrate that the spatio-temporal model can predict the existence of obstacles at future time slots.