Spatial language for route-based humanoid robot navigation

摘要

A more natural interaction between humans and mobilerobots can be achieved if they have a common ground(Kiesler 2005). Natural language interfaces support suc-cessful styles of interaction between robots and users.Typical scenarios include the user who instructs a robot toperform actions in an environment, such as moving tolocations or manipulating objects. Thus, route descriptions,which are used to guide robots in executing navigationtasks in their surrounding environment, are important forrealizing effective human–robot interaction. To provide theroute in an effective way, the rules and sequences ofcommands should be expressed in a very concise way.Natural language uses symbols and syntactic rules tointeract with the robots, which possess knowledge repre-sentations at the symbolic level. On the other hand, spatialreasoning concerning verbal route descriptions is essentialfor humans as well as for mobile robots which navigate inunstructured environments. Such reasoning gives robotsthe ability to comprehend human-like spatial language,which provides the human users with an intuitive interfacethat is consistent with their innate spatial cognition (Lauriaet al. 2001). It can also provide an accelerated learning byusing symbolic communication, which was proven byCangelosi and Hamad 2001.