Building a multi-resolution map for autonomous mobile robot navigation in living environments

This paper presents a method for building a multi-resolution map of a living environment to achieve efficient navigation. The standard practice for developing a representative map is to use an occupancy grid. However, the usage frequency of each environment has not been considered when building these environmental maps. These maps are not able to present the history of human activity as a feature of the environment. Therefore, these maps cannot be applied to mobile robot navigation to ensure safe and efficient robot movement in living environments. Therefore, we set out to design an environmental map containing resolution information of areas based on the movement history of humans and objects. For example, to present the position of static objects precisely, the area is required to be described in high resolution. An experiment was conducted to evaluate the proposed method and examine whether each resolution is changed appropriately based on the usage of the region.