This paper presents a technique which can reduce disturbance noise without an error microphone but by detecting Back Electromotive Force (BEF) of the canceling noise speaker. The error signal for active control is calculated in a program of a digital signal processor (DSP). Considerable noise reduction from 5dB up to 10 dB on sound evaluation microphone was obtained by this proposed method.

View full abstract

This paper describes a realization method of high kinetic energy utilizing mechanical elasticity within the joint moveable range on a multiple-joint robot. By utilizing mechanical elasticity, a robot can obtain high kinetic energy compared with a rigid robot. In this paper, we propose the "feedback excitation control", which is a control method to realize high kinetic energy utilizing elastic joints. Robot motion has to be kept within the joint moveable ranges. We propose a control method based on the "excitation limit hypersurface" to realize robot motions within the joint moveable ranges.

View full abstract

Many jumping robots have been studied, and many jumps by humanoid robots have been reported. This paper describes realization of jumping by a humanoid robot with spring equipment. The purpose of this work is to enable a low-power humanoid robot to jump. We have designed spring equipment and humanoid motions for jumping in consideration of performance of the humanoid's actuators. The humanoid can store energy in the spring by vibrating the spring with resonance frequency. Finally, the low-power humanoid could have temporarily higher energy with the elasticity of the spring, and we have achieved jumping by NAO that has its battery and CPU onboard.

View full abstract

This paper presents a modeling method for household objects for object fetching. Grasps for household objects are selected according to the intended use of the object. It is necessary to consider the functionality of the object for modeling. Our method, first divides the body of a household object into one or more parts in terms of its handling function. The functional parts are connected to each other with a link which has a joint attribute according to the parts configuration. The method offers an intuitive model for object fetching.

View full abstract

The service with an IC device spreads rapidly now, and IC device is becoming indispensable to our life. The IC student ID card was introduced in Kitakyushu National College of Technology. This student ID card has adopted the VRICS, and can utilize multiple services safely on a single card. VRICS is the new social information infrastructure technology to solve the problem of the current authentication system. As the service that takes advantage of student ID card, electronic-locks have been installed in school facilities. Along with installation of electronic-locks, system is desired that it can be managed all electronic-locks from a remote location. In this study, I develop the remotely and intensive management system of electronic-lock.

View full abstract

Now, service using an IC card increases quickly and IC cards are indispensable to our life. However, since the expenses of introduction or maintenance management including new issue of IC cards are high, for small and medium-sized enterprises, introduction is difficult for the actual condition. Then, IC card authentication system VRICS was developed for the purpose of enabling use of all the services with one card. In this research, in order to develop the service which can be used with the IC student identification card, the physical access control system which uses a robot gate and a student identification card is developed. Moreover, cooperation with service of others, such as an electronic lock, is also aimed at.

View full abstract

A new helicopter-type UAV has been developed in our research, which has stators just below the main rotor, instead of the tail rotor, to compensate the counter torque caused by the main rotor. In this paper, the blade number of the stators was reduced from 8 to 4 by optimizing the blade configuration and the control rigs were newly designed and manufactured in order to reduce the weight. Performance tests using ground test apparatus proved that the newly designed stators can compensate the counter torque at the same stator angle of attachment (25 degree) in the case that the rotor pitch angle is more than 7 degree. The maximum compensation torque was obtained at the angle of 40 degree. Moreover, the total thrust was hardly affected by the existence of stators.

View full abstract

This paper describes our development of small indoor Unmanned Aerial Vehicle (UAV) for the 7^<th> Student Indoor Flying Robot Contest. We designed a conventional airplane type UAV which is capable of dropping object to a designated point by hovering vertically. For this purpose, an attitude-control algorithm and a control system were developed. In order to estimate attitude quaternion from on-board accelerometer and gyroscope output, Extended Kalman Filter is used as the attitude-control algorithm. Then, in order to keep hovering using its control surface, the control system was designed and implemented with a microcontroller. The flight test revealed that the UAV is capable of vertical hovering in order to perform the missions required in the contest.

View full abstract

Aim of this research is development of a Micro unmanned vehicle (MUAV) named "Ptera Pod" to survey Mt. Mihara volcanic field in Izu-Oshima. These are assumed that agencies or volcano researchers to use at the time of the eruption. The fact is that the volcanic field have very strong wind, so that it is hard for operator to carry, assemble and fly MUAV in this situation. Ptera Pod is folding wing plane with one-dimensional mulch-joint redundancy mechanism, and also it is flying wing form. It is declared by field test in Ura-Sabaku of Izu-Oshima, these features make it easy, fast and safe to prepare the flight and control under strong wind condition around 10m/s. From the results, we confirmed that the flight onboard camera got HD movies about surfaces around the field successfully.

View full abstract

In this paper, we derive a mathematical model of a co-axial rotor unmanned helicopter, and we analyze the motion of the helicopter by using the derived model. First, we derive equations of motion of the helicopter by using velocity transformation method which is one of the multi-body dynamics modeling techniques. Forces and moments generated by the rotor are derived based on the aerodynamics of the rotor. Then, we verify the model by comparing the flight experiment data with model output. Finally, the motion analysis using the model is conducted to establish the behavior of the helicopter.

View full abstract

Quad rotor helicopters has simple mechanism and high stability. However, it is not easy for a quad rotor helicopter to fly long time and long distance, because most of the thrust is consumed to lift the body, and hence horizontal component of the thrust is small. In this paper, we present the experimental results of a quad rotor tail-sitter UAV (Unmanned Aerial Vehicle) which is composed of a quad rotors and fixed wing. Developed UAV can hover like a quad rotor helicopter, and can fly long distance like a fixed wing airplane. We designe attitude and altitude control system of the UAV for hovering. In order to verify the designed control systems, we conducted hovering control flight experiment. Additionally, we identified both the impact of propeller slipstream on yaw control and the solutions.

View full abstract

This paper presents attitude stabilization control of the tiltrotor aerial robot that we are developing for information gathering in the disaster area. Tiltrotor aerial robots are suitable to gather information rapidly and correctly, because they can achieve not only high speed and long-range flight but also stable hovering. In our tiltrotor aerial robot, a pair of coaxial rotors with tilt mechanism is mounted in the fuselage for the purposes of keeping the size of the robot small and reducing the risk that rotating rotors hit something. To stabilize the attitude of the robot, three ducted fans are also installed in the fuselage. We derive a model for the aerial robot in hovering flight and construct a control law to stabilize the attitude. Experimental results show that stable hovering flight can be successfully achieved by using the control law.

View full abstract

The authors developed a disk type flying robot capable of stable flight and a kite or balloon type flying robot capable of flight in the high winds. These flying robots have a radio communication device, a camera, sensor system for temperature, humidity, atmosphere, distance, position and environmental pollutant, and can provide the observed data to remote area. These data are helpful to draw up a hazard map.

View full abstract

As a system of aerial vehicle and landing gear for landing smoothly on general undulating ground, a prototype robot with quad helicopter and three leg mechanisms that has symmetric parallel links was developed. There are touch sensors and whisker sensors for each foot to detect the touchdown. Flying performance loading every control, actuation and power equipments and landing performance maintaining its body horizontal on stepped ground were evaluated experimentally.

View full abstract

The goal of this study is to develop a control method for levitation stabilization of an aerodynamic levitated train named the "Aero-Train," which is a high-speed, high-efficiency train system. The levitation in this system occurs because of the wing-in-ground (WIG) effect that acts on a U-shaped guideway. In order to achieve the goal of this study, authors have been studying on the control method of WIG effect vehicle. In the previous report, the modeling, controller design and simulation results of the Aero-Train prototype along Y and Z axes and about roll, pitch and yaw axes have been illustrated. In this paper, development of a small prototype of Aero-Train and experimental results which confirmed the effectiveness of designed controller are described.

View full abstract

This paper proposes a dependable visual control method for a small-scale helicopter with a wireless camera. The camera sometimes captures noisy images, and we cannot control the helicopter by directly using the captured images. The 3D position, the posture and noise information are given by a parallel template matching technique. The proposed method does not require any known markers.

View full abstract

This paper proposes autonomous flight control system of RC helicopter with wind disturbance. Autonomous RC helicopter is widely adopted in many fields such as surveillance, intelligence and rescue operations. Especially, it is expected that the super small-size RC helicopter is adopted in various environment. However, the low thrust and the light weight of the super small-size RC helicopter make it difficult to control the RC helicopter autonomously with wind disturbance. The relation between maximum takeoff weight of each helicopter and their corresponding difficulty of control with disturbance in previous works is given in this paper. To solve this problem, we present control law and conduct experiments for hovering flight with wind disturbance.

View full abstract

We developed a centipede-like multi-legged robot with passive intersegment joints, named Mukade 1. This robot is controlled based on Follow-the-Contact-Point (FCP) gait control. The purposes of development of this robot are verification of realizability of the FCP gait control on a real robot, and acquisition of elemental technologies to realize a centipede-like multi-legged robot. We realized real-time control of Mukade1 via 8 MHz and 8 bit PIC microcomputers network. Finally, the ability of Mukade1 was verified through the experiments of walking of even and uneven terrains.

View full abstract

This paper proposes a novel method, based on formal verification, to automatically derive the parameters of a centipede-like multi-legged robot which is controlled based on Follow-the-Contact-Point (FCP) gait control. Centipede-like multi-legged walking robots are capable of making diverse motions, but the parameters concerned of the walk control and the mechanical configuration are difficult to be derived analytically because of the redundant degrees of freedom. Our proposed method is based on repetition of applying formal verification to each parameters set. We modelled the behavior of a centipedelike multi-legged robot based on the FCP gait control and derived ambulatory specifications for it. Finally, in order to show the usefulness, we plotted two parameters extracted from the various parameters which were derived from the proposed method.

View full abstract

Flocking algorithm of multi agent system is robust and disaster tolerant even though agents face obstacles or destructed. In this paper, we focus on applying these mechanism to single humanoid robot by assuming that the robot is the set of agents. We extend flocking algorithm to be capable of using in single robot and apply it to humanoid robot to maintain standing posture.

View full abstract

This paper addresses the self-configuration problem for a swarm of autonomous mobile robots. As our solution approach, an adaptive local interaction algorithm is proposed to allow individual robots to form different equilateral triangular configurations according to their local distributions. Specifically, Delaunay triangulation is applied to calculate the areas of individual triangles generated around each robot. From the computation, the proposed algorithm computes the average area. Next, each robot estimates a side length enabling to form an equilateral triangle with the same average area. By using the proposed algorithm, robot swarms can self-configure themselves while adapting to their distribution conditions. Through extensive simulations, we verify the effectiveness of the proposed algorithm.

View full abstract

With the ability to handle critical but dangerous tasks instead of human, robot swarms have been regarded as one of the emerging fields in robotics research. One of the major topics is how to locate various types of sources with swarms of mobile robots. This paper presents a simple yet efficient decentralized algorithm for swarm mobile robots to localize odor sources based on local odor intensity. Swarm mobile robots with generally limited sensing range and communication capability relocate themselves independently based on the gradient of local odor intensity and coverage of the area within their limited sensing range. Simulation results show that the proposed decentralized algorithm enables swarms of mobile robots to localize distributed odor sources efficiently even in situation where multiple odor sources exist. This algorithm does not only provide simplicity but also efficiency, as it requires less computation cost and improves the convergence time significantly compared to the Voronoi-based algorithm.

View full abstract

We propose a two-dimensional position and angle sensors using cheap the steel-mark. The proposed mark can cover all zones in path width. Therefore, the mobile robot can be driving at an intermittent setting. This mark can also detect steel plate with in long liftoff with robustness by using the magnet for environment changes. We have developed a sensor mechanism and signal processing system to detect the landmark. The sensor was installed to the mobile robot. It showed useful for robot applicable.

View full abstract

We thought method to reduce error of GPS data. This method is how to reset acquisition data at regular interval using the landmark constructed in the road. But this method is not able to remove mutation noise. So we made location estimation by filter with least square of fourth-order approximation. In addition, mobile robot prospect of advance in running stability by overlap a self-location of GPS and dead reckoning.

View full abstract

In this research, we propose a mobile robot localization based on GPS error evaluation in urban environments. Localization in urban environment has still a challenging problems such as multipath and signal lost. Our proposed method uses a localization with GPS which is evaluated the reliability of GPS observations, dead reckoning and integration method using Divided Difference Filter (DDF). Proposed evaluation algorithm uses satellites elevation and azimuth that are leveraged in global positioning system in order to calculate covariance matrix. Especially, our evaluation focus on calculated real-time deviation from covariance matrix. The covariance matrix is associated with DDF. Hence, the proposed system makes mobile robot localization less likely to be affected by accuracy of GPS observations. The effectiveness of the proposed method is proved through experiments in urban environments.

View full abstract

This paper describes an improvement for the accuracy of GPS single point positioning using L1-SAIF signal and carrier smoothing. L1-SAIF is a performance enhancement function of Quasi-Zenith Satellites System (QZSS). Carrier smoothing is the method for smoothing a pseudorange by precise carrier phase. The proposed technique mitigates systematic and accidental errors of GPS positioning. But, carrier smoothing has two problems. One is increasing the positioning error in long-time smoothing. Another is vulnerability to cycle slip. So, in this paper, we propose the method for decreasing the effects of these problems using L1-SAIF. The evaluation results confirm the effectiveness of the proposed technique and the feasibility of its highly accurate positioning.

View full abstract

This paper describes a localization method for vehicles using GPS. To reject multipath error, the method discern GPS satellite visibility and remove signals from invisible GPS satellites using 3D-Map. GPS antenna position on the 3D-Map is estimated by using a particle filter. Each particle has different visibility of GPS satellites. Therefore, every particle has different positioning solutions. The method is useful for localization and navigation of vehicles between buildings.

View full abstract

We are developing the automatic guided vehicles for the outdoors. and studyingabout the robust localization method using a laser scanner for the stable autonomous run. Then, the influence which the climate condition containing snowfall has on observation of a laser scanner was investigated.

View full abstract

This paper presents a mathematical model of a mobile modular robot that spontaneously switches locomotion patterns, inspired by plasmodium of true slime mold. Significant features of this robot are threefold: (1) each module of the robot has a truly soft and deformable body, stemming from elastic elements, filled with fluid, (2) these several identical modules are physically connected with fluid-filled tubes, which induces long-distance interaction between the modules; and (3) by exploiting the physical interaction between the body parts, a fully decentralized control using phase oscillators with completely local sensory feedback mechanism is realized. The preliminary simulation results show that altering softness of the body part according to the amount of attractant allows the robot to change locomotion direction spontaneously in a fully decentralized manner. The results obtained are expected to shed new light on the design scheme for life like robots that exhibit versatile, agile and adaptive behavior.

View full abstract

Although autonomous decentralized control is a key concept for the realization of robots that can adapt to various evironments in real time, a systematic way of designing it is still lacking. We tackle this problem through the development of a two-dimensional sheet-like robot that can obtain rich sensory information. Recently we extended the curvature derivative control of the snake-like robot to a two-dimensional bodily structure, and showed that the simulated sheet-like robot could locomote on irregular terrain; however, environment to which it can adapt was limited. In this paper, we add a local sensory feedback mechanism that uses local pressure on the body surface to the previous control scheme to increase adaptability to environments. Simulation results show that the sheet-like robot can locomote effectively on unstructured terrain by exploiting scaffolds. We also show a physical robot now developing.

View full abstract

Autonomous decentralized control is a key concept for building robots capable of exhibiting adaptive behavior under unpredictable real world constraints. However, it is still unclear how to generate non-trivial macroscopic functionalities of an entire system using only locally available information. In this study, we focus on the slithering locomotion of snake-like robot as a case study, and show an answer to this problem based on variational principle and reaction-diffution equation.

View full abstract

Snakes activity utilize terrains irregularities and attain propulsion force by pushing their body against scaffolds. Our objective is to clarify the mechanism underlying this locomotion. We have previously proposed an autonomous decentralized control scheme in which the curvature derivative control method is combined with a reflexive mechanism using local pressure on the body wall, and have confirmed its validity via simulation. In this study, we develop a physical robot to investigate the validity of the proposed control scheme in the real world.

View full abstract

Snakes actively utilize terrain irregularities and attain propulsion force by pushing their bodies against scaffolds. The objective of this study is to understand the mechanism underlying this locomotion. Our previous control scheme of a snake-like robot, in which the curvature derivative control method is combined with a reflexive mechanism using local pressures on the body wall, could not fully reproduce innate behavior of real snakes. In this study, we improve our previous control scheme. We show that the locomotion of a simulated robot with the improved control scheme is surprisingly in good agreement with the locomotion of a real snake.

View full abstract

We have been developed an autonomous running method for a mobile robot with a SOKUIKI sensor by a laser which can be run on an outdoor road in uncertain 3-D world. In this paper, we propose a path-planning and an autonomous running method to direct this robot to goal at the shortest route using a made 2-D map by a bird's-eye measurement with the SOKUIKI sensor.

View full abstract

In this study, we aim to realize an autonomous mobile robot which works for a car. One of the essential functions to achieve this purpose is estimation of the robot position relative to the car. We propose a new estimation method of the relative position of the robot by using the car shape measured with the laser range finder.

View full abstract

In recent years, the demand for autonomous mobile robot in outdoor environment has been increasing. In order to guide the robot, it is essential to get the current position. In this study, we propose a method to correct the self-position of the robot, which is estimated by odometry, by aerial imagery and SOKUIKI sensor. First, we explore how to superimpose the reflection point cloud on top of the aerial imagery. Next, matching the edge data of aerial imagery and the upper edge of the building from the inclined reflection point cloud corrects the self-correct of the robot, by which the robot was able to run on the straight road about 60m autonomously.

View full abstract

In this paper, we propose a method for detecting and tracking moving obstacles in an outdoor environment. Our system clusters ranges observed by a laser range-finder. Moving obstacles are not estimated from clusters directly, we assume clusters as candidates of moving obstacles. The system recognizes moving obstacles in the candidates to track them with Extended Kalman Filter. The effectiveness of the proposed method is proved through experiments in outdoor environments.

View full abstract

In this paper, we propose a method to track moving obstacles in unknown environment. Unlike controlled environment such as that to use experiment, robots might collide with moving objects such as other moving robots and people under real environment, and thus it is important that the robot can track moving obstacles for avoiding them. In conventional research, moving objects have been tracked for removing them as outliers before executing SLAM. In contrast, we employ the SLAM, which can be applied to dynamic crowded environments, and then track moving objects using particle filter. We show the result of applying the proposed method to real environment, and discuss the effectiveness of the method.

View full abstract

This research aims to develop the mapping of a wheeled inverted pendulum type robot. In this paper, initial work on the generation of a 3D map of an indoor environment using a 3D laser range finder and occupancy grid mapping is presented. From the generated map the structure of the indoor environment was represented. It is suggested that the 3D laser range finder can be used to generate a precise 3D map of an indoor environment.

View full abstract

In this study, we developed multiple mobile robots to perform group escape behavior by mimicking fish school. A simulation system and an experiment system are developed for robots evaluation. The actual value of parameters of robots and control algorithms are used in the simulation system. For observing the group motion behavior, a software system is developed for processing the images from celling cameras to obtain positions and attitudes of robots in the experiment system. With the simulating and experimental system, basic group behavior of robots developed has been evaluated.

View full abstract

This paper addresses a new cooperative theory of multiple robots which adapts them to environmental changes. In practical applications with multiple robots, the environmental elements, which are target object condition, number of working robots, and so on, can change easily through their interactions. The new cooperative theory adapting environmental changes is needed. The proposed cooperative theory is inspired by social cooperative characteristics of human-beings, especially subjective evaluation and imitation. This work is experimented by Sheep-Dogs Simulation (SDS) to validity of the theory. The SDS consists of multiple sheep-dogs, named dog-agents, and multiple sheep-agents. In conclusion effectiveness of the theory was proposed in SDS. Multiple robots would be adapted to environmental changes by the theory.

View full abstract

Ants act in a group of many colleagues and efficiently using the pheromone. We want to achieve that motion in the robot. Therefore, must express the pheromone. In this paper, we represent the pheromone with the paper that changes color when exposed to water. We made a robot that meets the criteria and did the experiment. As a result, we achieved some of the pheromone trail.

View full abstract

The purpose of this study is to generate the map to estimate the robot position by using stereo cameras and odometry. When creating a map image of the feature points is calculated using the Speeded Up Robust Feature(SURF). Matching of feature points obtained from the cameras using Fast Library for Approximate Nearest Neighbors(FLANN). Accuracy of the map is raised by using the Distance-Based Outlier. And processing time of image processing can significantly reduce the by the use of Graphics Processing Unit(GPU). Self-localization experiments were performed using the map. The robot position and orientation can be estimated with the stereo cameras and odometry.

View full abstract

This paper describes a localization technique for a mobile robot in an environment with many unknown obstacles, such as pedestrians. To realize robust localization against unknown obstacles by using a particle filter, a space observation model has conventionally been used. Although this conventional model is very effective, it only uses two-dimensional information, which is dependent on the scan plane of laser scanner. If a model is extended to three dimensions, the observed information increases, and the the accuracy of the position estimation improves in comparison with that of the two-dimensional model. In this research, the proposed model extends the conventional model to three dimensions. Therefore, the proposed model still has the advantages of the previous model; namely, the likelihood of a particle can be evaluated by a simple method using a particle filter, and the occupied space can be considered. In a real environment, it was shown that the proposed model is more effective than the two-dimensional conventional model.

View full abstract

For autonomous navigation of a mobile robot, 3D map has some advantages over 2D map particularly in dynamic environments where occlusion may occur due to moving objects. This paper discusses the expression of the 3D map for estimation of planar position and pose. We propose a single layer expression where the most significant cuboid (MSC) is recorded on each cell of the 2D grid. Experimental comparison with MLS map show there is no significant difference in navigation performance.

View full abstract

We developed an acquisition method of maps and paths using ROS stacks for autonomous navigation by teaching playback framework. There are many strategies for autonomous navigation. In teaching playback framework, it is not a problem that maps are distorted continuously. However, if the maps are accurate, it is useful for human interfaces and extending the autonomous navigation area. Recently, the ROS platform has been used actively around the world. In ROS stacks, there is slam_gmapping stack that provides Grid-based SLAM using Rao-Blackwellized Particle Filter. Therefore we are trying to use ROS stacks for acquisition of maps and paths, and to evaluate its effectiveness for autonomous navigation.

View full abstract

Estimating a robot's own position is an essential requirement for autonomous mobile robots. Visual Odometry (VO) can play a key role in navigation tasks in slippery natural terrain, which highly degrade the accuracy of Wheel Odometry, without relying on other infrastructures nor any prior knowledge. A challenge of VO system lies in how to increase the number of features tracked between continuous frames in untextured terrain. Although there has been many feature selection algorithms proposed, none of them are sufficient in terms of calculation speed and robustness. Hence, this paper propose an algorithm that dynamically switches between high-speed/low-accuracy and low-speed/high-accuracy algorithm according to the texture of terrain. Validity of this algorithm was proved by the simulation using datasets taken at Izu-Oshima, Japan.

View full abstract

In extreme environments, such as lunar surface or disaster area in the earth, it is expected for mobile robots to explore instead of human. However, in such environments, wheels or tracks of the robots are typically slipped, and thus, localization based on odometry is unreliable. To improve odometry localization, our research group has been developing a non-contact translation measurement device that uses a red laser module and an optical device. In this research, we developed a localization module that consists of the measurement device, Inertial Measurement Unit (IMU), and Global Positioning System (GPS). Furthermore, we developed 3D range system based on the localization module and a Laser Range Finder (LRF). In this paper, we introduce the localization module and an initial experimental result of recognition of 3D environment.

View full abstract

In premise wiring, we insert the leading wire into a pipe to install a communication line. However, some buildings don't have pipes. So we have proposed an inchworm mobile robot using suction pads and pneumatic actuators to install a communication line without pipes. Additionally, this robot need to incorporate arm mechanisms to control the move direction and avoid obstacles. In this paper, we report the fundamental evaluation of an arm mechanism and an inchworm mobile robot.

View full abstract

In recent years, some UGV have been developed that run over irregular terrain with remote control. They have a number of DOF to improve their mobility and are capable of running over stairs and steps. However, their remote operations are complicated, because they are controlled by operators using a control pad such as used in computer games or toys. In order to easily control robots with many DOF, we propose an intuitive remote control system, using the conventional master-slave operation. Our UGV 'KUROGANE' consists of a typical crawler plus a human-like upper body. KUROGANE has achieved the ability to negotiate swiftly over irregular terrain. We demonstrated performance over outdoor rugged terrain as well as steps and stairs. In this paper, we investigated the effectiveness of the performance by simulation.

View full abstract