Arnau Carrera

Towards Autonomous Robotic Valve Turning

Abstract In this paper an autonomous intervention robotic task to learn the skill of grasping and turning a valve is described. To resolve this challenge a set of different techniques are proposed, each one realizing a specific task and sending information to the others in a Hardware-In-Loop (HIL) simulation. To improve the estimation of the valve position, an Extended Kalman Filter is designed. Also to learn the trajectory to follow with the robotic arm, Imitation Learning approach is used. In addition, to perform safely the task a fuzzy system is developed which generates appropriate decisions. Although the achievement of this task will be used in an Autonomous Underwater Vehicle, for the first step this idea has been tested in a laboratory environment with an available robot and a sensor.

Toward persistent autonomous intervention in a subsea panel

Intervention autonomous underwater vehicles (I-AUVs) have the potential to open new avenues for the maintenance and monitoring of offshore subsea facilities in a cost-effective way. However, this requires challenging intervention operations to be carried out persistently, thus minimizing human supervision and ensuring a reliable vehicle behaviour under unexpected perturbances and failures. This paper describes a system to perform autonomous intervention—in particular valve-turning—using the concept of persistent autonomy. To achieve this goal, we build a framework that integrates different disciplines, involving mechatronics, localization, control, machine learning and planning techniques, bearing in mind robustness in the implementation of all of them. We present experiments in a water tank, conducted with Girona 500 I-AUV in the context of a multiple intervention mission. Results show how the vehicle sets several valve panel configurations throughout the experiment while handling different errors, either spontaneous or induced. Finally, we report the insights gained from our experience and we discuss the main aspects that must be matured and refined in order to promote the future development of intervention autonomous vehicles that can operate, persistently, in subsea facilities.

Online Discovery of AUV Control Policies to Overcome Thruster Failures

We investigate methods to improve fault-tolerance of Autonomous Underwater Vehicles (AUVs) to increase their reliability and persistent autonomy. We propose a learning-based approach that is able to discover new control policies to overcome thruster failures as they happen. The proposed approach is a model-based direct policy search that learns on an on-board simulated model of the AUV. The model is adapted to a new condition when a fault is detected and isolated. Since the approach generates an optimal trajectory, the learned fault-tolerant policy is able to navigate the AUV towards a specified target with minimum cost. Finally, the learned policy is executed on the real robot in a closed-loop using the state feedback of the AUV. Unlike most existing methods which rely on the redundancy of thrusters, our approach is also applicable when the AUV becomes under-actuated in the presence of a fault. To validate the feasibility and efficiency of the presented approach, we evaluate it with three learning algorithms and three policy representations with increasing complexity. The proposed method is tested on a real AUV, Girona500.

An Intervention-AUV learns how to perform an underwater valve turning

Intervention autonomous underwater vehicles (I-AUVs) are a promising platform to perform intervention task in underwater environments, replacing current methods like remotely operate underwater vehicles (ROVs) and manned sub-mersibles that are more expensive. This article proposes a complete system including all the necessary elements to perform a valve turning task using an I-AUV. The knowledge of an operator to perform the task is transmitted to an I-AUV by a learning by demonstration (LbD) algorithm. The algorithm learns the trajectory of the vehicle and the end-effector to accomplish the valve turning. The method has shown its feasibility in a controlled environment repeating the learned task with different valves and configurations.

AUV based multi-vehicle collaboration: Salinity studies in Mar Menor coastal lagoon

Abstract An experiment with different AUVs was carried out in the Mar Menor Coastal Lagoon from October 31 to November 5 in order to measure and assess the influence of the water from the Mar Menor on the adjacent area of the Mediterranean. This was carried out as a result of the meeting held between several institutions from the Iberian Peninsula and EEUU (see Vilanova Marine Science/Robotics Meeting 2010). The experiment was to launch several AUVs at the same time in different zones of the Mediterranean and Mar Menor lagoon. AUVs took salinity data trying to do a coordinated mission during two operative days (November 3 and November 5). Others days of the experiment were used to the vehicles preparation and error correction (November 2 and November 4). This paper presents the steps followed in preparation and operative days with the set of AUVs. This paper presents also the salinity results obtained during these missions.

Sonar-based chain following using an autonomous underwater vehicle

Tracking an underwater chain using an autonomous vehicle can be a first step towards more efficient solutions for cleaning and inspecting mooring chains. We propose to use a forward looking sonar as a primary perception sensor to enable the vehicle operation in limited visibility conditions and overcome the turbidity arisen during marine growth removal. Despite its advantages, working with acoustic imagery raises additional challenges to the involved image processing and control methodologies. In this paper we present a robust framework to perform chain following, combining perception, planning and control disciplines. We first introduce a detection system that exploits the sonars high frame rate and applies local pattern matching to handle the complexity of detecting link chains in acoustic images. Then, a planning system deals with the dispersed detections and determines the link waypoints that the vehicle should reach. Finally, the vehicle is guided through these waypoints using a high level controller that has been tailored to simultaneously traverse the chain and keep track of upcoming links. Experiments on real data demonstrate the capability of autonomously follow a chain with sufficient accuracy to perform subsequent cleaning or inspection tasks.

Cognitive system for autonomous underwater intervention

Learning by demonstration (LbD) has been adapted to the underwater domain.LbD has been extended to control simultaneously an AUV and its manipulator.An underwater valve turning task has been used to test the LbD algorithm.The autonomously reproduced task achieves similar performance as human operator.The method proves to be resilient to perturbations. The implementation of autonomous intervention tasks with underwater vehicles is a non-trivial issue due to the challenging and dynamic conditions of the underwater medium (e.g., water current perturbations, water visibility). Likewise, it requires a significant programming effort each time that the vehicle must perform a different manipulation operation. In this paper we propose, instead, to use a cognitive system that learns the intervention task from an expert operator through an intuitive learning by demonstration (LbD) algorithm. Taking as an input few operator demonstrations, the algorithm generalizes the task knowledge into a model and is able to control the vehicle and the manipulator simultaneously to reproduce the task, thus conferring a more adaptive behavior in front of the environment changes and allowing to easily transfer the knowledge of new tasks. A cognitive architecture has been implemented in order to integrate the LbD algorithm with the onboard sensors and actuators and to allow its interplay with the vehicle perception, control and navigation modules. To validate the full framework we present real experiments in a water tank using an AUV equipped with a four DoF manipulator. A human operator teaches the system to perform a valve turning intervention and we analyze the results of multiple task reproductions, including cases under the effect of water current perturbations, showing the success of the system in autonomously reproducing the task.

Adaptive frequency filtering for forward-looking sonar imagery spectral registration

In the last few years, forward-looking sonar devices have emerged as a powerful perception alternative for those underwater environments with reduced visibility. Thanks to its capability to deliver high quality acoustic images at a near-video frame rate, they can be regarded as the analogous tool of optical cameras for operations conducted in turbid waters. However, despite the analogy, the particularities of forward-looking sonar imagery pose a significant challenge to the techniques typically used on optical images and, especially, to the key step of image registration, essential in applications like mosaicing, sonar-aided navigation or image denoising. In this sense, previous investigations have encouraged the use of spectral registration methods as a promising alternative over the traditional feature-based registration approaches used on optical images. In this paper, we propose to improve the spectral registration of forward-looking sonar images with an adaptive filtering technique that allows to cope with the noise and variability inherent to the forward-looking sonar image registration problem. Results show that by using the proposed filtering we achieve a more accurate pairwise alignment of the sonar images that can benefit subsequent processing in many applications.

Intervention Payload for Valve Turning with an AUV

This paper presents an intervention payload for an AUV working in a valve turning operation in free-floating control mode. The payload consists of a stereo camera for panel detection, a 4 degrees of freedom electrical manipulator and a specifically designed end-effector, which contains a force and torque sensor, an in-hand camera and a passive effector for valve operation. This payload was designed to be integrated in Girona 500 AUV in the context of an oil application, in which a valve panel must be operated by turning some of the T-bar handles. The paper describes the design of the payload and its interaction with AUV. It also describes the perception systems that have been developed to detect and operate the valves. Experiments in a water tank show the performance of the AUV and the suitability of the payload.