Researchers from the national project RAUVI [Reconfigurable autonomous underwater vehicle for intervention], which is coordinated by Pedro Sanz, a lecturer of Computer Science and Artificial Intelligence at the UJI, have successfully tested the autonomy of the robot for developing underwater intervention tasks. The robot has managed to recover an object similar to an aircraft black box without the direction of any operator. All these experiments have formed the basis for making a demonstration before experts of the European Commission during the annual review of TRIDENT European project, which is also coordinated by the Universitat Jaume I.
The test was performed a few weeks ago at the Universitat de Girona, where there is a pool suitable for experimentation on underwater robotics. During the meeting, the researchers also tested the performance of the three parts involved in the experiment: the robotic arm, which is being improved by the UJI; the vehicle, in which the Universitat de Girona is working, and the computer vision techniques, which are being developed by the Universitat de les Illes Balears.
In the first part of the experiment the vehicle in which the robot was anchored descended to the bottom of the pool to survey the area using computer vision techniques and to draw a map. After that, the researchers asked the robot to recover an object (a black box), and the vehicle with the robotic arm plunged again, sought the object with the required characteristics, picked it up and pulled it to the surface.
The research group Interactive and Robotics Systems (IRS Lab) is composed by Prats, Juan Carlos García, José Javier Fernández and Raúl Marín and is led by Pedro Sanz. The team obtained the first results of manipulation underwater just two weeks ago, playing grip simulations with the robotic arm in a water tank installed in an office at the UJI before going to Girona to make the joint integration testing. The arm has four articulations, two at the shoulders, one at the elbow and the fourth at the wrist. It can also open its hand, which is claw-like but has T-shaped slots that enable to anchor on it cables or tools for picking up objects.
Achieving such a project would reduce the economic and human resources efforts which are attached to submarine operations, since support vessels or umbilical cables would not be necessary, nor ROV pilots responsible for teleoperation in conditions that involve fatigue and stress. Thus, this point would enable to carry out operations which would be impossible for teleoperated systems and which require a continuous connection through an umbilical cable to a support ship, a characteristic that affects the vehicle dynamics and limits the travel distance of the robot.
At present, the first application which more than 40 researchers are seeking is the recovery of black boxes with this autonomous action system, but some other potential application scenarios that could benefit from this project would be certain tasks associated with marine biology, performing routine practices such as taking samples (eg rock, water or sand); at permanent observatories, in lifesaving, in health care tasks to the diving teams such as lighting in a particular area, or assistance in the use of some kind of tool, to name only a few examples.
The project is funded with 530 000 euros by the Spanish Ministry of Science and Innovation, within the VI National Plan for Scientific Research, Development and Technological Innovation 2008-2011. Each one of the participating universities is responsible for a specific subproject. The UJI is the responsible for the mechanical integration of the robotic arm, the visual tracking of the object of interest, the control of the arm for handling tasks and the interface to specify these tasks.
The Universitat de Girona is the responsible for generating the navigation and mechatronics systems of the underwater vehicle where the robot is docked. The Universitat de les Illes Balears is the responsible for assisting in the planning and guiding of the movements necessary to achieve the autonomous navigation of the robot, using advanced computer vision techniques.
Cite This Page: