Personal Intelligent Vehicle that Extends Human Physically and Informationally

点击翻译按钮获取中文摘要

We have been conducting a research project of a personal intelligent vehicle called AT(Attentive Townvehicle). AT has a function of automatic driving to the destination that its passenger specified.One of example applications of AT is a museum appreciation support system that enriches museum experiences for each user who is a beginner of art appreciations. The system can encourage effective museum experiences to users by personalizing appreciation courses based on their past museum experiences. The system can tacitly record the museum experiences of the users who move in museums by boarding the automated vehicles.AT is capable of omni-directional movement and rotation by using special wheels called mecanum wheels and four servo motors. AT also has a function of automatic avoidance of obstacles by sensing 360 degrees around AT using four laser range sensors when its passenger specified a destination to automatically move.During automatic driving, it is difficult for AT to avoid moving o … More bstacles that are approaching AT from dead areas of AT's sensors. Therefore, we solved the problem by extending sensing areas of AT based on a Small Unmanned Vehicle(SUV). SUV has a function of generation of a 2D environmental map for localization of current position. SUV can search for places where there are some dead areas of AT on the way to its destination. SUV autonomously moves to the places and probes circumstances around there as a substitute of AT. Based on the above the mechanism, We developed a method of automatic safe transportation and evaluated its underlying techniques.As another useful application of SUV, we developed a mechanism to automatically generate indoor 3D maps by using a SUV. Our method can make a precise alignment between the 3D maps and the 2D environmental maps so that our generated 3D maps can be applied to highly informative indoor navigations.In addition, based on AT's basic functionality, we also developed a hand gesture interface as an intuitive control interface of AT. Hand shapes and motions are recognized by a depth sensor in real time. We evaluated our gesture interface from the viewpoint of visual feedback of recognition processes and found that our interface is intuitive enough for novice users to control AT as easy as possible.The outputs of this research can contribute to facilitation of R & D and deployment of intelligent moving machines and realization of safe and comfortable society in which senior citizens and handicapped people can live independently. Less