We report the first experimental results of the navigation method of multiple Autonomous Underwater Vehicles (AUVs) for wide seafloor surveys proposed by us in our previous reports. The key idea of the proposed method is that moving AUVs estimate their states (horizontal position and heading angle) based on AUVs remaining stationary on the seafloor (landmark AUVs), alternating between these roles to expand the observational coverage. Moving AUVs land on the seafloor and transmit compressed information about their estimated states to the landmark AUVs when they change the roles. Sea trials were carried out using two AUVs to evaluate the performance of this method. One is the AUV Tri-Dog 1 (TD) which can move and estimate the states and the other is a Dummy AUV (DA) which cannot move but can estimate the states. In the experiments, TD and DA alternatively estimated the states. To evaluate the positioning accuracy of the proposed method, the states were recalculated based on sensor data obtained at sea trials in post processing. As a result, positioning errors of the proposed method were estimated to be smaller than dead reckoning. The Errors were found to be about 0.4 m in horizontal position and about 1.5 degree in heading angle. The method demonstrated to be successful in stable positioning with AUVs alternating between the moving role and the landmark role. Future works include the verification of the method using two AUVs in sea environment to realize wide seafloor survey.

Optimal and quasi-optimal navigations of an AUV achieved by the minimum time guidance in undersea area of sea currents are presented. When an AUV is to transfer to a given destination within an area of sea current disturbance, ingenious guidance enables the minimization of navigation time. In this study, a numerical procedure to obtain solution of the optimal guidance law in undersea area of sea current disturbances is presented. Simulated optimal and quasi-optimal navigations by minimum time guidance are realized based on the dynamics of an AUV ldquor2D4rdquo. The r2D4 is a deep ocean exploring AUV developed by Institute of Industrial Science, the University of Tokyo. The developed algorithm never fails to derive the optimal headings for minimum time navigation, if only the distribution of current velocity in navigation region is deterministic.

In this paper, autonomous underwater vehicle (AUV) navigation schemes are proposed using forward looking sonar (FLS) and charged coupled device (CCD) camera for near bottom applications. Scans obtained from the onboard FLS are processed using a feature extraction technique discussed in the paper to extract stable point features in the environment. A technique is also proposed for AUVs to track optical features. Through field trials conducted using the test-bed vehicles NTU-UAV and the Twin-Burger2, the performance of the algorithms proposed are analyzed. The paper discusses the above-mentioned individual sensor capabilities and drawbacks in AUV navigation.

Autonomous underwater vehicles (AUVs) are suitable for condition survey of the surface of artificial structures such as pillars and caissons in harbors. This paper introduces a method for AUVs to trace artificial structures using a light-section ranging system. This system acquires the profile of the target objects over a wide area using a sheet laser beam. The vehicle navigates by referencing the principal shape of the structure, and traces its surface while taking video images of it. This ranging system also enables 3D mapping of its surroundings based on high accuracy positioning. The method is implemented on the testbed AUV "Tri-Dog 1" and its performance is verified through tank tests. It is shown that the AUV navigates robustly against small obstacles and floating particles. Precise depth mapping of the tank floor is also carried out using the same system.

Autonomous underwater vehicles (AUVs) have great advantages for activities in deep sea, and expected as the attractive tool. However, AUVs have various problems which should be solved. In this paper, the self-organizing map (SOM) is applied as the clustering method for the navigation system. The SOM is known as one of the effective methods to extract the principle feature from many parameters and decrease the dimension of parameters. Through the competitive learning algorithms, the obtained map is tuned to express specific features of the input signals. We have been investigating the possibility of a navigation system based on SOM through simulations and experiments with an AUV called "twin-burger". The learning algorithm of usual SOM is unsupervised learning. However, supervised learning algorithms should be introduced because the relationship between distances information and desirable behavior of the robot, that is, the relationship from inputs to outputs should be acquired and learned. In this paper, a supervised learning algorithm is introduced into SOM and a method to adapt the local map to its environment by learning and evaluating the trajectory of robot is proposed. In the proposed method, the "initial map" is made static and digital vale as teaching data. In order to include more information of environment in the initial map, the trajectories of robot are evaluated, and the evaluation is utilized in the learning process. This method enables the map to have both the effect of dynamics of robot and environmental information. The efficiency of the method is investigated through the simulations and experiments.

Autonomous underwater vehicles (AUV) are suitable for condition survey of artificial structures such as pillars and caissons in harbors. This paper describes a method to trace the structure's surface using a light-section profiling system. This profiling system determines the continuous shape of the target objects over a wide area by the light sectioning method. The vehicle navigates referencing the principal shape of the structure, tracing its surface while taking video images. This method also enables three-dimensional mapping of traced structures and the seabed. The method is implemented using testbed AUV "Tri-Dog 1" and verified by tank tests. The AUV navigates robustly against trivial objects such as small obstacles and floating particles. Precise depth mapping of the tank bottom is carried out.

Underwater vehicles are expected as the attractive tools for the operation in the extreme environment such as the deep ocean survey. In order to realize the useful and practical robots, which can work in the ocean, underwater vehicles should take their action by judging the changing condition from their own sensors and actuators, and are desirable to make their behavior with limited efforts of the operators, because of features caused by the working environment. Therefore, the mobile robot should be autonomous and adaptive to their environment. Development of the path planning system, which can navigate the vehicle without the collision to the obstacles, is one of the most important problems in order to realize the autonomous underwater vehicles. In this paper, Self-Organizing Map (SOM) proposed by Kohonen is applied to the navigation, which takes the distances to the surroundings as inputs, and outputs the direction for the robot to proceed. The efficiency of path planning system is investigated through the simulation and experiments with an underwater vehicle "Twin-Burger".

A training algorithm for motion planning of a nonholonomic autonomous underwater vehicle (AUV) in the strong water current is proposed in this paper. The proposed algorithm can be applied in the environment with obstacles placed in arbitrary configuration. In order to realize these functions, the Q-learning and teaching method are introduced and a multilayer structure is proposed. By introducing Q-learning, the motion of the nonholonomic AUV can be suitably treated. Taking advantage of the Baysian net, a motion planning algorithm in the case of an existence of obstacles, is derived automatically from the learned knowledge. The multilayer structure accelerates the learning process. Results of the demonstration by the simulation of control of R-One robot show the high performance of proposed algorithm.

To observe underwater objects such as artificial piles and natural chimneys by taking visual images, autonomous underwater vehicles (AUVs) have advantages that they have no problem related to the umbilical cable which may get entangled with the target and surrounding obstacles. This paper proposes a navigation method for AUVs to observe such objects based on images by a CCD camera with simple laser pointers, on the assumption that the visibility is enough for taking their visual images. It includes a localization method under a distorted magnetic field. The proposed method was implemented in the developed testbed AUV "Tri-Dog 1", and proved experimentally through tank tests. The vehicle accomplished the observing mission that automatically detects three pile-objects and turns around them which were modeled on actual piles in a harbor. Based on the proposed method, AUVs are able to carry out a survey of the target object in detail without any help from outside.

This paper proposes a fast reinforcement learning (RL) algorithm for motion planning algorithm of nonholonomic autonomous underwater vehicle (AUV) in the strong water current with high reliability. The proposed algorithm can also be applied in an underwater environment with obstacles which are placed in arbitrary configuration. The algorithm has the ability to complete a learning process within a practical time limit. In order to accomplish a high learning speed, this paper proposes a hierarchical RL algorithm that copes with the curse of dimensionality, which comes from high complexity of dynamics of an AUV. The higher level of the algorithm refers only the position of the AUV, and learns a motion planning algorithm. The lower level of the algorithm refers to the velocity of the AUV and compensates for the undesirable nonMarkovian effects by stabilizing the dynamics motion of the AUV. The proposed hierarchical algorithm also uses a dynamics model described in the form of Bayesian-network. The dynamics model is very useful for increasing the efficiency of searching in learning process. The proposed algorithm is demonstrated by simulation. The result of the simulation shows high learning speed and performance of the proposed algorithm.

In this paper, a sensor fusion technique is proposed for autonomous underwater vehicles (AUV) to track underwater cables. The paper discusses the navigation of AUV when cable is invisible in the image, and the selection of the correct cable (interested feature) when there are many similar features appearing in the image. The proposed sensor fusion scheme uses deadreckoning position uncertainty with a 2D position model of the cable to predict the region of interest in the image. This reduces the processing data increasing processing speed and avoids tracking other similar features appearing in the image. The proposed method uses a 2D position model of the cable for AUV navigation when the cable features are invisible in the predicted region. An experiment is conducted to test the performance of the proposed system using the AUV "Twin-Burger 2". The experimental results presented in this paper shows how the proposed method handles the above mentioned practical problems.

A sensor fusion technique is proposed for autonomous underwater vehicles (AUVs) to track underwater cables. The focus of this paper is to solve the two practical problems encountered in optical vision based systems in underwater environments: (1) navigation of AUV when cable is invisible in the image; and (2) selection of the correct cable when there are many similar features appearing in the image. The proposed sensor fusion scheme uses the dead reckoning position uncertainty with a 2D position model of the cable to predict the region of interest in the image. This reduces the processing data, increases the processing speed and avoids tracking other similar features appearing in the optical image. The proposed method uses a priori map of the cable for AUV navigation when the cable features are invisible in the predicted region in the image. An experiment was conducted to test the performance of the proposed algorithm using the AUV "Twin-Burger 2". The experimental results obtained show how the proposed method handles the above-mentioned practical problems.

A training algorithm for dynamics control of nonholonomic AUV (autonomous underwater vehicle) is proposed in this paper which can recover from thruster failure during cruising mission. It is based on Q-learning and teaching method. The back up data that represents dynamics model expressed in the form of Bayesian net can be used effectively in this case. In order to overcome difficulties due to, making discrete expression of continuous state space of AUV, the algorithm uses multiresolution Q-value tables which is combined in the form of subsumption architecture. Simulation results show high performance of the proposed algorithm for a vertical ascent mission in a severe current condition. It is shown that AUV users can conveniently and quickly train the control algorithm of the AUV by using simulation of dynamics of the vehicle.

In this paper, a sensor fusion technique is proposed for autonomous underwater vehicles (AUV) to track underwater cables. The focus of this paper is to solve the two practical problems encountered in optical vision based systems in underwater environments; namely 1) navigation of AUV when cable is invisible in the image, and 2) selection of the correct cable (interested feature) when there are many similar features appearing in the image. The proposed sensor fusion scheme uses dead reckoning position uncertainty with a 2D position model of the cable to predict the region of interest in the image. This reduces the processing data increasing the processing speed and avoids tracking other similar features appearing in the optical image. The proposed method uses a priori map of the cable for AUV navigation when the cable features are invisible in the predicted region in the image. An experiment is conducted to test the performance of the proposed algorithm using the AUV "Twin-Burger 2". The experimental results presented in this paper show how the proposed method handles the above-mentioned practical problems.

In this paper, a sensor fusion technique is proposed for autonomous underwater vehicles (AUV) to track underwater cables. The work presented here is an extension of the authors' previously proposed vision based cable tracking system (1997, 1998). The focus of this paper is to solve the two practical problems encountered in vision based systems; namely (1) navigation of AUV when cable is invisible in the image, and (2) selection of the correct cable (interested feature) when there are many similar features appearing in the image. The proposed sensor fusion scheme uses deadreckoning position uncertainty with a 2D position model of the cable to predict the region of interest in the image. This reduces the processing data increasing processing speed and avoids tracking other similar features appearing in the image. The proposed method uses a 2D position model of the cable for AUV navigation when the cable features are invisible in the predicted region. An experiment is conducted to test the performance of the proposed system using the AUV "Twin-Burger 2". The experimental results presented in this paper shows how the proposed method handles the above mentioned practical problems.

Describes a neural network system which executes identification of robot dynamics and controller adaptation in parallel with robot control. The system consists of two parts: real-world part and imaginary-world part. The real-world part is a feedback control system for the actual robot. In the imaginary-world part, the model of robot and the controller are adjusted continuously in order to deal with the change of dynamic property caused by disturbance and so on. The system is designed to be suitable for a computer system with parallel processing ability. In the paper, adaptability of the controller system is investigated by heading keeping and path following experiments on the condition that unknown disturbances are given to the robot.

This paper describes the design and overall structure of a test-bed system for developing multiple autonomous underwater robot systems. The developed system can be regarded as a multi-agent system which consists of a pair of autonomous underwater robots, named "The Twin-Burger-I and II", an ultrasonic positioning system, and a human diver. An ultrasonic command link system is used for communication among each agents. The distributed vehicle management architecture (DVMA) is applied to the Twin-Burger robot as an architecture for the control software of the robot. Mission execution experiments are carried out in the Institute of Industrial Science (IIS) testing pool. The results show the effectiveness and applicability of the DVMA.

A vision based system is developed for interaction between the man-robot unit and robot-robot unit in underwater environments. Vision is the communicating medium between the agents. The image consists of an ON/OFF light pattern produced by a set of electro-luminescent panels, representing a particular command. The vision system introduced in this paper recognizes these patterns under the defined environmental conditions. In the image processing system the Hough transform is performed with the Sobel operator to extract important features of the image. The decision making time of this system is approximately 1 sec on a transputer based hardware, which is equivalent to the CPU system on the testbed robot, "The Twin-Burger". This decision making time is acceptable for slowly moving underwater robots. In order to track the region of interest, the vision system calculates the pan and tilt angles of the CCD camera. The system is tested for different situations underwater as well as in air.