Acoustic motion estimation and guidance for unmanned underwater vehicles

The problem of navigation, guidance and control of Unmanned Underwater Vehicles (UUVs) is addressed in this paper. A task-function based guidance system and an acoustic motion estimation module have been integrated with a conventional UUV autopilot within a two-layered hierarchical architecture for closed-loop control. The design of the guidance system is based on suitable Lyapunov functions that can handle the different manoeuvres involved in approaching a target. Range and bearing information provided by a pencil beam profiling sonar are processed by an Extended Kalman Filter based algorithm for motion estimation in a structured environment. The resulting Navigation Guidance and Control (NGC) system has been tested on Roby2, the UUV testbed developed at the Istituto Automazione Navale of Italy's National Research Council. The experimental set-up as well as modalities and results are discussed.     

Experimental study on fine motion control of underwater robots

This paper considers thruster dead zones and saturation limits, which are nonlinear elements that complicate fine motion control of underwater robots. If the vehicle is configured with redundant thrusters, the respective dead zones and their surrounding nonlinear regions could be avoided by implementing a null motion solution for the command input of the vehicle. This solution is derived from the vehicle's geometry and is realized before the application of the motion control algorithm. The result is an improvement in system performance exclusive of the implemented controller type. The approach is illustrated through simulation and experiment with an underwater robot, ODIN.     

Real-time optical SLAM-based mosaicking for unmanned underwater vehicles

This article discusses the possibility of building in real-time a mosaic of the seafloor relying on a simultaneous localization and mapping (SLAM) framework. The goal is to provide an unmanned underwater vehicle with a relatively rough visual map of the seafloor to support basic navigation and context awareness. To achieve that goal, an accurate estimation of the location of the visual landmarks and, in particular, the correct data association when a visual landmark is re-visited by the vehicle are the crucial points. Instead of using a global mosaic, this work uses the combination of a set of local mosaics constructed in the vicinity of the SLAM visual landmarks. The contributions of this article are mainly the use of SURF features, the local mosaics approach and the real-time capability. The use of SURF features allows eliminating false positives in the data association of SLAM visual landmarks. The local mosaics approach is an effective way of correcting the effects of the drift on the mosaic in real time. The main contribution is the real-time capability as it will be seen. The algorithm was tested using a batch of experimental data in typical operating conditions and the results prove the effectiveness of the approach.     

Multi-Sensory Motion Estimation and Control of an Autonomous Quadrotor

In this paper, a fully autonomous quadrotor in a heterogeneous air–ground multi-robot system is established only using minimal on-board sensors: a monocular camera and inertial measurement units (IMUs). Efficient pose and motion estimation is proposed and optimized. A continuous-discrete extended Kalman filter is applied, in which the high-frequency IMU data drive the prediction, while the estimates are corrected by the accurate and steady vision data. A high-frequency fusion at 100 Hz is achieved. Moreover, time delay analysis and data synchronizations are conducted to further improve the pose/motion estimation of the quadrotor. The complete on-board implementation of sensor data processing and control algorithms reduces the influence of data transfer time delay, enables autonomous task accomplishment and extends the work space. Higher pose estimation accuracy and smaller control errors compared to the standard works are achieved in real-time hovering and tracking experiments.     

On the identification of non-linear models of unmanned underwater vehicles

This documentation presents a comparison between two identification methods for the off-line identification of non-linear models of unmanned underwater vehicles (UUVs), one based on the minimization of the acceleration prediction error (direct method) and another one based on the minimization of the velocity one step prediction error (integral method). The direct method has already been used in UUV's identification (IFAC Conference CAMs’ 2001, Control Application in Marine Systems, Glasgow, Scotland, UK, 2001). Our new proposal, the integral method, can be applied to a quite general class of non-linear multivariable models and is characterized by an excellent numerical performance. Both methods are compared through their application to the identification of the dynamic model of URIS UUV. Results suggest that better models can be obtained using the proposed method (integral method).     

Exact non-linear Bayesian parameter estimation for autonomous compliant motion

This paper presents theoretical and experimental results for the estimation of large position and orientation inaccuracies during force-controlled compliant motion. This is a significant improvement over previous results. The estimation is based on position, velocity and force measurements. For large position and orientation inaccuracies, the non-linear estimation problem is not satisfactorily solved by existing Kalman filters. Therefore, a new Bayesian estimator is derived. The filter is derived independently of our application and is valid for static systems (parameter estimation) with any kind of non-linear measurement equation subject to Gaussian measurement uncertainty and for a limited class of dynamic systems. Experimental results for the estimation of the inaccurately known positions and orientations of contacting objects during autonomous compliant motion are presented.     

The bio-inspired model based hybrid sliding-mode tracking control for unmanned underwater vehicles

In this paper a novel hybrid control strategy is developed for trajectory tracking control of unmanned underwater vehicle (UUV). The proposed hybrid control strategy consists of two subsystems: a virtual velocity controller and a sliding-mode controller. The tracking errors are shown to asymptotically converge to zero by Lyapunov stability theory using the new approach, whereas in the traditional backstepping method, speed jump occurs if the tracking error changes suddenly. The biologically inspired model is designed to smooth the virtual velocity controller output, avoid speed jumps of underwater vehicles and satisfy the thruster control constraint. The effectiveness and efficiency of the proposed control strategy are demonstrated through simulations and comparison studies.     

Design of a vision-guided microrobotic colonoscopy system

In this paper presents a novel design of a microrobotic colonoscopy (MRC) system. The proposed microrobotic colonoscope is an autonomous vision-guided device, which is designed to navigate inside a human colon for the purpose of observation, analysis and diagnosis. It is developed to alleviate the shortcomings in the existing manual colonoscopy procedure, which is generally cumbersome and tedious for the colonoscopist and painful to the patients. The MRC system is divided into three areas, i.e. design of the microrobotic device, path planning and guidance, and offboard control system. A novel design of the microrobot is presented which utilizes a pneumatic mechanism to achieve locomotion and steering. A new concept of clamping the colon wall based on passive vacuum devices is suggested. General mathematical analysis governing the differential steering of the robotic tip is also described. The path planning of the microrobot is carried out based on the sensory fusion utilizing the quantitative parameters derived from the captured images and the tactile sensors. An off-board control system to control the directional movements of the microrobot is explained. The proposed colonoscopy system was tested with physical models and animal colons, and the experimental observations are presented.     

Image-based predictive display for high d.o.f. uncalibrated tele-manipulation using affine and intensity subspace models

The addition of immediate but estimated visual feedback, called predictive display, improves tele-manipulaton performance when the real video feedback is delayed. Current systems typically rely upon a previously calibrated camera and manipulator. We present a method where the motor-visual calibration is estimated on-line from motor commands and returned video images only. Predicted visual feedback is presented in two forms. As soon as a basic model has been estimated, a wire frame drawing of the predicted current pose is overlaid on the delayed video feedback. After some time when a rich model has been estimated, predicted intensity images are synthesized and these replace the delayed real video. In an intermediate situation where blurry synthesized images can be computed, the wireframe is overlaid on the synthesized images to show precisely the pose of the object. Experiments with a Utah/MIT robot hand and PUMA robot arm are shown.     

Visual Odometry for a Hopping Rover on an Asteroid Surface using Multiple Monocular Cameras

Visual odometry refers to the use of images to estimate the motion of a mobile robot. Real-time systems have already been demonstrated for terrestrial robotic vehicles, while a near real-time system has been successfully used on the Mars Exploration Rovers for planetary exploration. In this paper, we adapt this method to estimate the motion of a hopping rover on an asteroid surface. Due to the limited stereo depth resolution and the continuous rotational motion on a hopping rover, we propose to use a system of multiple monocular cameras. We describe how the scale of the scene observed by different cameras without overlapping views can be transferred between the cameras, allowing us to reconstruct a single continuous trajectory from multiple image sequences. We describe the implementation of our algorithm and its performance under simulation using rendered images.     

Final-state control of a two-link cat robot

In this study, we deal with the twisting motion of a falling cat robot by means of two torque inputs around her waist. The cat robot consists of two rigid columns and has two internal actuators at the joint to generate torque inputs around normal coordinates. This system is a nonholonomic system whose angular momentum is conserved. We formulate the state equation that has torque inputs to the joint by using the nonholonomic constraint and the Lagrange-d'Alembert principle. Then, we transform the system into a linear parameter varying system. In order to improve error learning of a final-state control method, we provide the initial inputs in order to determine the appropriate rotation direction in the early stage of the twisting motion. Next, we introduce the method of the artificial potential function to the final-state control in order to make the maximum bending angle small. The feedforward torque inputs can be obtained by the final-state control in order to bring the system from the initial state to the final state in the desired time. In simulations, we also demonstrate that the twolink cat robot can land on her feet by using the 2-d.o.f. control system even when her waist damping coefficient varies.     

Vision-guided flight stability and control for micro air vehicles

Substantial progress has been made recently towards designing, building and test-flying remotely piloted micro air vehicles (MAVs) and small unmanned air vehicles. We seek to complement this progress in overcoming the aerodynamic obstacles to flight at very small scales with a visionguided flight stability and autonomy system, based on a robust horizon detection algorithm. In this paper, we first motivate the use of computer vision for MAV autonomy, arguing that given current sensor technology, vision may be the only practical approach to the problem. We then describe our statistical vision-based horizon detection algorithm, which has been demonstrated at 30 Hz with over 99.9% correct horizon identification. Next, we develop robust schemes for the detection of extreme MAV attitudes, where no horizon is visible, and for the detection of horizon estimation errors, due to external factors such as video transmission noise. Finally, we discuss our feedback controller for selfstabilized flight and report results on vision-based autonomous flights of duration exceeding 10 min. We conclude with an overview of our on-going and future MAV-related research.     

Visual control of two aerial micro-robots by insect-based autopilots

In the framework of our research on biologically inspired microrobotics, we have developed two novel Automatic Flight Control Systems (AFCs): OCTAVE ( Optical flow Control sysTem for Aerospace VEhicles) and OSCAR ( Optical Scanning sensor for the Control of Autonomous Robots), both based on insects' visuomotor control systems. OCTAVE confers upon a tethered aerial robot (the OCTAVE robot) the ability to perform terrain following. OSCAR gives a tethered aerial robot (the OSCAR robot) the ability to fixate and track a contrasting target with a high level of accuracy. Both OCTAVE and OSCAR robots are based on optical velocity sensors, the principle of which is based on the results of previous electrophysiological studies on the fly's Elementary Motion Detectors (EMDs) performed at our laboratory. Both processing systems described are light enough to be mounted on-board Micro-Air Vehicles (MAVs) with an avionic payload small enough to be expressed in grams rather than kilograms.     

Parallel Implementations of Block-Based Motion Vector Estimation for Video Compression on Four Parallel Processing Systems

Parallel algorithms, based on a distributed memory machine model, for an exhaustive search technique for motion vector estimation in video compression are being designed and evaluated. Results from the execution on a 16,384 processor MasPar MP-1 (an SIMD machine), a 140 node Intel Paragon XP/S and a 16 node IBM SP2 (two M IMD machines), and the 16 processor PASM prototype (a partitionable SIMD/MIMD mixed-mode machine) are presented. The trade-offs of using different modes of parallelism (SIMD, SPMD, and mixed-mode) and different data partitioning schemes (the rectangular and stripe subimage methods) are examined. The analytical and experimental results shown in this application study will help practitioners to predict and contrast the performance of different approaches to parallel implementation of this important video compression technique. The results presented are also applicable to a large class of image and video processing tasks. Case studies, such as the one presented here, are a necessary step in developing software tools for mapping an application task onto a single parallel machine and for mapping a set of independent application tasks, or the subtasks of a single application task, onto a heterogeneous suite of parallel machines.     

Force and acceleration sensor fusion for compliant manipulation control in 6 degrees of freedom

This paper focusses on sensor fusion in robotic manipulation: six-dimensional (6-D) force/torque signals and 6-D acceleration signals are used to extract forces and torques caused by inertia. As result, only forces and torques established by environmental contact(s) remain. Apart from an improvement of hybrid force/pose control behavior, an additional major benefit is that regular resetting/zeroing of force/torque sensors before free space/contact transitions can be omitted. All essential equations, transformations and calculations that are required for this 6-D fusion approach are derived. To highlight the meaning for practical implementations, numerous experiments with a six-joint Staeubli RX60 industrial manipulator are presented and the achieved results are discussed.     

A study on a brachiation controller for a multi-locomotion robot — realization of smooth,continuous brachiation

In this paper, we present a control method to realize smooth continuous brachiation. The target brachiation is basically divided into two actions: a swing action and a locomotion action. In order to realize the continuous brachiation effectively and smoothly, it is necessary to start the swing action as soon as the robot grasps the front target bar at the end of the locomotion action. The collision, which occurs at the moment the robot grips the target bar, affects the pendulum motion of the robot. The action of bending the elbow joint of the swinging arm is proposed in order to solve this gripping problem. The elbow-bending action enables the robot to decrease the impact forces and use the excess mechanical energy after the end of the locomotion phase. Thus, there is no loss of energy and waste of time during the subsequent swing phase. Experimental results show that the robot can successfully achieve smooth, continuous brachiation.     

Coordination of Multiple Control Schemes for Mobile Robot Navigation on the Basis of the Generalized Stochastic Petri-Nets

There have been two major streams of research for the motion control of mobile robots: model-based deliberate control and sensor-based reactive control. Since the two schemes have complementary advantages and disadvantages, each cannot completely replace the other. There are a variety of environmental conditions that affect the performance of navigation. The motivation of this study is that multiple motion control schemes are required to survive in dynamic real environments. In this paper, we exploit two discrete motion controllers for mobile robots. One is the deliberate trajectory tracking controller and the other is the reactive dynamic window approach. We propose the selective coordination of two controllers on the basis of the generalized stochastic Petri net (GSPN) framework. The major scope of this paper is to clarify the advantage of the proposed controller based on the coordination of multiple controllers from the results of quantitative performance comparison among motion controllers. For quantitative comparison, both simulations and experiments in dynamic environments were carried out. In addition, it is shown that navigation experiences are accumulated in the GSPN formalism. The performance of navigation service can be significantly improved owing to the automatically stored experiences.     

Compliant motion using a mobile manipulator: an operational space formulation approach to aircraft canopy polishing

The operational space formulation provides a framework for the analysis and control of robotic systems with respect to interactions with their environments. In this paper, we discuss its implementation on a mobile manipulator programmed to polish an aircraft canopy with a curved surface of unknown geometry. The polishing task requires the robot to apply a specified normal force on the canopy surface while simultaneously performing a compliant motion keeping the surface of the grinding tool tangentially in contact with the workpiece. A human operator controls the mobile base via a joystick to guide the polishing tool to desired areas on the canopy surface, effectively increasing the mobile manipulator's reachable workspace. The results demonstrate the efficacy of compliant motion and force regulation based on the operational space formulation for robots performing tasks in unknown environments with robustness towards base motion disturbances. The mobile manipulator consists of a PUMA 560 arm mounted on top of a Nomad XR4000 mobile base. Implementation issues are discussed and experimental results are shown.