System Design for the Next Generation Subsea IMR Vehicles
MetadataVis full innførsel
This thesis is about how to unlock the full potential of unmanned underwater vehicles to save cost and increase quality in inspection maintenance and repair (IMR) services for subsea petroleum extraction equipment. Today IMR all operations are performed with a ship as the base. An initial analysis of the IMR tasks that are performed today indicate that more than one third of all operations can be performed without the ship. At the same time the ship is the main cost driver of such operations. To be able to perform IMR operations without a ship there is a need for a new type of underwater vehicle. These next generation IMR vehicles need to be capable of operat- ing independently of a ship. The thesis explores concepts for such vehicles, and analyzes strengths and weaknesses of different concepts. Seven possible technology gaps is identified. These gaps are: • Human machine interface and autonomy • Reliability • Field integration • Communication • Power • Standardization • Navigation The industry is currently working on closing the gaps with field integration, communic- ation, standardization and power. The main area of research in this thesis in the areas of human machine interface, autonomy as well as underwater navigation. An open-source simulation environment for underwater vehicles and robots is developed to make research and development of navigation and control systems faster and easier. The simulation environment allows the user to simulate underwater robotic vehicles with realistic dynamic behavior in a 3-dimensional virtual environment. The environment is highly configurable, and offers a set of modules for simulating different types of vehicles in a number of underwater scenarios. The simulator can be used for control system de- velopment, path planning, risk management and testing in a safe virtual environment. The possibility for virtual testing will lower the cost and reduce time spent on operations. The simulation environment enables the user to make a virtual replica of a specific underwater robot system. Such replica may be used as a software-in-the-loop system for testing and verification of control systems and algorithms. A study on automated module handling and installation is done to explore the capabilities of automated underwater vehicles. Subsea modules often need to interface with existing equipment when they are installed on the seafloor. The modules will therefore need to be positioned and oriented precise. Today this is often done by using guide-wires, or by using a manually controlled ROV. Both these methods are costly and time consuming. The study propose a system for automating the ROV. It is believed that such system will make underwater vehicles capable of automated module installations. For the next generation IMR vehicles it is proposed to use complementary sensor techno- logies to get both close range precision navigation and long range navigation for transits. In this thesis an effort has been made to handle time delays in the acoustic measurements to increase the real-time performance of such navigation solution. As a complimentary technology to acoustic navigation, a vision based underwater localization system for underwater vehicles using fiducial markers is proposed. The system is using an Extended Kalman Filter to fuse visual information and inertial information from an IMU-sensor to a 6 degree of freedom pose estimate. The work describes the development of a Kalman Filter for this application and includes testing of the filter on an underwater vehicle in a pool. A high accuracy underwater motion capture system in the pool is used as ground truth for performance assessment of the proposed system. The main objective has been to develop a visual localization system for underwater vehicles performing automated manipulation at subsea facilities. Experimental validation of the proposed filter confirms that it is a good candidate for high precision localization of underwater robots performing autonomous manipulation work. The next generation IMR vehicles need the ability to both operate with a large degree of autonomy and to be able to let humans control and supervise critical and hard operations. The thesis addresses the challenge of combining automated control with having a human operator in the loop. This is done by using theory Human Centered Automation in other industries. The theory is adapted and extended to work fur the subsea IMR scenario. This resulted in a prototype of a human centered control system for IMR operations. The control system has been tested both in simulation and using small remotely operated underwater vehicles in a pool. The overall goal of the thesis has been to take a systems perspective in the development of the next generation of subsea IMR services. This has been done by contributing to a better understanding of the problem, and by taking the technologies where gaps was identified a step further. There is however still research and development left before the full potential of next generation IMR vehicles can be unlocked.