项目针对度量地图导航方法人机交互不友好、不自然、占用计算资源大等问题，以空间认知学为理论依据，融合SLAM技术、物体检测技术、自然语言处理技术，构建面向人机共融环境的室内移动机器人空间语义地图。在此基础上，依据景物的空间语义关系，探索一种高效稳健的室内定位技术；依据人体工学、空间距离学和可供性理论，实现导航目标点的自主选择；依据空间语义信息，建立机器人社交空间模型，研究机器人与人和物品友好的路径规划技术。本项目具有多学科交叉融合的特点，研究成果有望实现人-机-物共融的导航技术。这种将人的环境认知框架引入机器人环境描述的方法有理论价值，对促进机器人融入人的生活场景具有实用价值。

To address the issues that the navigation technologies based on the metric map are unfriendly, less natural and large resources occupancy, this project presents the concept of Spatial Semantic Map based on the theory of spatial cognition for indoor mobile robots in human-robot coexistence environments. The Spatial Semantic Map will be built up by integrating the SLAM technology, object detection technology and natural language processing technology. On the basis of the Spatial Semantic Map, an efficient and robust localization technology will be explored according to the spatial relations of the scenes and objects. Benefiting from the results of ergonomics, proxemics and affordance theory, the navigation target points will be selected independently and human friendly. A human-robot-object friendly path planning technology will be explored by introducing a social space model for indoor robots. The project features a cross-disciplinary research and its results are expected to achieve a harmonious human-robot-object navigation strategy. The methods of this project are theoretically valuable by introducing the spatial cognition frame of human into the robotic environment description, and are practically valuable in integrating service robots into human-life scenes.