Active Indoor Mapper

*Active Mapping for Indoors Mobile Robot*

— We propose an active mapping strategy to actively build an indoors map by a mobile robot that is undergoing a self-exploration mode. Our work consists of real-time 6D pose estimation, active loop closure, pose correction by numerical optimation, and mapping building. In the following two maps, the left one is  created by a human operator remotely controlling the robot, while the right one is the result by the robot itself according to our method. Our active mapping can exploit a variety of single sensors, e.g., a single-line lidar or an RGB-D camera, or a combination of a couple of sensors, cheers!

Reference paper: 

Zezhou Sun, Banghe Wu, Cheng-Zhong Xu, Sanjay Sarma, Jian Yang, and Hui Kong, Frontier Detection and Reachability Analysis for Efficient 2D Graph-SLAM Based Active Exploration, IEEE International Conference on Intelligent Robotics and Systems (IROS) 2020 (oral presentation)