In this paper, we develop a model of the behavior of visitors in an exhibition space, and design an amenity space to reduce congestion. Because the flow of visitors depends on the space layout, it is possible to reduce the congestion by changing the layout of exhibits. To deal with the congestion, a macro model of visitors is required. The flow of visitors is modeled by two-dimensional vector field and the individual behavior is represented by the dynamic model including collision avoidance of individuals. To model walking and stopping behaviors, we introduce an extended dimensional space where the extra coordinate represents the viewing time of each exhibition. The parameters in the proposed model are identified from the measured data of human flow line. To reduce the congestion, the layout of exhibits is optimized based on the proposed model by minimizing the collision avoidance vector. The proposed method is verified by simulations and experiments using swarm robots, which consist of autonomous mobile robots and radio controlled robots.
Recently, many developed countries have begun to experience the problems associated with an aging society. One problem in particular is the growing number of older adults in the labor who suffer from various physical restrictions. Aiming to develop technology to assist these individuals, the novel power assisted system is proposed in this paper. This system estimates the force applied to the object from both the ground reaction force and the human behavior based on the human body dynamics model with 13 links. Then, estimated force is used for the power assistance to help the conveyance of the heavy object. The validity of the model is verified through both simulation and experiment with three tasks by four subjects, and it is shown that our proposed system can be used as a power assisted system.
This paper describes a robot perception of an unknown object for service robots using range camera. There are plane detection methods like RANSAC and Hough Transform to detect planes of unknown objects. However, the previous method has problems which are high computational costs and low-accuracy of small object detection. We propose the plane detection method based on retinal structure to solve aforementioned problems. The proposed method is constructed by simple plane detection based on retinal structure, and integrated object plane detection. The simple plane detection based on retinal structure is focused on small plane detection and reducing computational costs. The integrated object plane detection is focused on stability of detecting plane on the specific object. As experimental results, we show that the computational cost of proposed method is no more than 10% compared to previous plane detection method. And, the proposed method detects small planes of specific object. Furthermore we discuss the capability of proposed method which coordinate the ability of reducing computational costs and improving the plane detection accuracy.
This paper describes development of a navigation system for moving through narrow spaces. Getting a path through narrow spaces is a difficult task for wheeled robots with non-holonomic constraints because it is necessary for the path planning process to carefully detect collision considering contour of robot and realize the smoothness of the path. In addition, in order to generate accurate paths, the path planning process needs an accurate map which shows the obstacles even on narrow slopes. Therefore we proposed the methods for generating obstacle maps with the plane link structure approach and optimizing path using nonlinear optimization approach. By using our methods, collision free smooth path through narrow spaces with slope has been generated in real time and our robot is able to move narrow spaces.
We describe a gaze stabilization based on fixation in robotic vision. Robot vision system in this study consists of CMOS cameras, gyro sensor, micro controller, and Field Programmable Gate Array (FPGA). During robot moves, the appropriate fixation point is selected and changed according with the robot position. The movement of two cameras and head is controlled based on the information obtained from cameras and gyro sensor. In addition to the camera movement, we apply an image shift so that the fixation point is fixed to the center of the view field. This method can fix the target point to the center of the view with high accuracy and respond to a fast and large movement of the robot.