Functionality is not the only major objective of service robots design, but safety and affinity are also necessary. This study focuses on a design policy for a service robot that can be used to transport people in a real world environment. In this paper, a design policy which referred to the rules of a real world robot challenge (RWRC) called “Tsukuba Challenge 2010” and an idea to design the robot's body which is suitable for navigation method based on internal sensors, are presented. The design policy emphasizes functionality, safety, and affinity. Therefore, a service robot which can ensure the functionality and the safety during coexistence with humans can be developed. In 2010, we developed a robot with affinity and applied it in many events including Tsukuba Challenge. A questionnaire on the robot's perfection level was taken during the events. Feasibility of the proposed policy is shown by the results and performance of the demonstrations.
This paper presents the development of autonomous mobile robot called “Orange 2010” that is designed for participating both Intelligent Ground Vehicle Competition 2010 and Tsukuba Challenge 2010. In order to adopt both competition and challenge rules, we newly build mobile robot based on electric wheelchair. The combination of omnidirectional camera and laser range scanner enables superior environment recognition especially at outdoor environment. Three different path planning algorithms are developed by using MATLAB language for each competition at IGVC2010 and Tsukuba Challenge 2010. Validity of a control software program developed for each challenge is examined by actual competitions.
This paper introduces our approach to RWRC (“Tsukuba Challenge”). We are developing “intelligent software modules” for mobile robots in “Intelligent RT Software Project.” “Intelligent software modules” have mobile robot basic functions for running in real world where people exist. RWRC is a challenge aiming to research and develop the mobile robots in the presence of human activity and it is a chance to evaluate our modules. We implemented ICP(Iterative Closest Point) scan matching with upper landmarks, explicit motion planing, collision and obstacle avoidance in our robot named “TUFS,” with our software modles. As a result, our robot archieved the goal of RWRC2009 and RWRC2010.
The Tsukuba Challenge is a proving ground for mobile robots, the task of which is autonomous navigatoin along a predefined path of 1km in outdoor environments. We joined the Tsukuba Challenge and achieved the tasks in 2009 and 2010. This paper focuses on the problem analysis and basic design of our navigation system, and presents the design policy and lessons learned in the experiments. Our system consists of gyro-assisted odometry, a roundly-swinging 3D laser scanner, a fish-eye camera, and a localization method using map matching and a particle filter. The integration of these technologies made our system highly reliable to achieve the task.
Autonomous outdoor navigation technology is required for automatic porterage, transportation, and many other service robot systems. In case of people, one can go through the town to the destinated building without using global positioning system. He walks while detecting his orientation and knowing which street he is on. And he finds the intersection and turns based on his knowledge of the map. In this paper, we explain our outdoor navigation method based on the road following with detecting orientation and finding intersection, which we adopted in our robot for Tsukuba Challenge 2010. We have used two scanning laser range sensors, for obstacle detection and finding the edge of the road, to know the road area, and a magnetic field sensor to know the robot orientation to follow the proper path. In this paper, we describe the implementation of the system, and show the experimental results of autonomous navigation in park and pedestrian street of Tsukuba city.
In this paper, we describe a robust and fast self position estimation technique for mobile robot in an environment where many unknown obstacles exist. The free-space observation model is the basis of the proposed technique. Although the free-space observation achieves robust self-position estimation, it has a complicated likelihood evaluation, and particles continue to spread to a direction where are no landmarks. In this research, we solve these problems with two likelihood evaluations based on the area of free space and occupied space. We evaluated the robustness and verify the validity of the proposed method by a simulation and an experiment in a real environment.
We are researching transporter mobility robots for a new generation of personal transportation infrastructure. This paper describes our autonomous navigation technology which is adapted to crowded pedestrian streets, and evaluation results in Tsukuba Challenge 2010 using an experimental robot Sofara-T. In crowded outdoor environments, there are many moving objects such as pedestrians and bicycles. And there are complex shaped objects such as trees and bushes with branches and leaves. For autonomous navigation, it is required that localization, road terrain analysis (obstacle detection) and local path planning (obstacle avoidance) methods are adaptable to dynamic environments and complex shaped landmarks. Primarily, we developed a 3D-LIDAR using gimbal mechanism to measure 3D shapes of surrounding objects as landmarks and obstacles at rapid speed. In localization, we utilize static objects in upper regions than humans as landmarks in order not to lose robot position caused by occlusion in crowds. Moreover we developed a new localization method with 3D map matching adapted to complex shaped landmarks. Our method extracts bounding directed points of occupied space as landmarks, and matches current measured points to reference map points utilizing normal directions and a constraint of the gravity direction. For highly reliable road terrain analysis, we fuse results of bump detection with a 3D-LIDAR and a stereo camera using a binary bayes filter. Then our local path planning in variable velocity can avoid obstacles with safety and smoothness. In Tsukuba Challenge 2010, our robot Sofara-T robustly ran 1.8[km] 14 times in 15 experiments at maximum velocity 3.96[km/h].
This paper proposes multi-sensor localization framework for an autonomous mobile robot using multiple sensor. To realize robust localization in outdoor environments, many external sensors are needed. The propoesd localization framework in the paper is to divide the localization system into modules that contains retro-active localization function for particle filter to solve computation delay of to calculate measurement models. In this paper, solutions for multi-sensor localization are illustrated and localization experimental results in outdoor environments are described.
Smart Dump has been developed to complete the mission of Real World Robot Challenge, or Tsukuba Challenge. The challenge requires robots to travel autonomously in pedestrian environment for 1[km] or more distance without any help of human operators. It is even prohibited to modify the environment for the robots. This paper reviews three-time participation to the challenge, especially unexpected results due to the real world.
This paper describes development and consideration of autonomous navigation based on tracking waypoints without the environmental information such as an accurate geometry map for a mobile robot. The waypoints consist of positions which are measured by the DGPS. It is known that some positions by DGPS are rough. Therefore, a free space extraction is necessary for stable navigation. The robot detects free space on the own front by a LRF, and autonomously tracks the waypoints. However, the robot sometimes encounters a complicated situation in which there are unknown obstacles or pedestrians in the actual environment, and misses the correct path. Therefore, a recovery behavior is necessary for achieving autonomous navigation. In Tsukuba Challenge 2010, although it was almost completed that the robot autonomously navigates to the goal, failures such as course out were sometimes occurred. The reason of every situation of the course out was verified by experiments in which the robot moves from wide space to narrow space. In this paper, performance of the waypoints tracking based navigation and its consideration in cases of complete and retired navigaitions in the experiments are shown.
Constructing environmental information is one of key problems for robots to extend their operation areas. To solve this problem, we introduce the data creation and mapping methodology in GIS field into environmental information construction for robots, in which lots of developers play their own rolls to provide geographical information, i.e. environmental information, for users. We attempted Tsukuba Challenge 2009 without any on-site trials nor measurements by our robot and constructed environmental information based on the public maps of Tsukuba Chuo Koen where Tsukuba Challenge 2009 was held. As a result, we completed the mission of the first round. This means that the methodology is effective for robots.