Autonomous Underwater Docking Using Active/Passive Electro-location

In the context of Remotly Operated Vehicule (ROV) for underwater application, connecting the robot to its docking station can be a significant challenge [2]. ROV have to navigate in a complex and unpredictable environment since the perception is limitted. Optical and Acoustic sensor can not be used is this kind of environment because blind spots can occur and short range obstacles may interfere. To perform the docking, it also necessary to provide it a guide. In those situations, with poor visibility, to increase the perception, we have to use an other sense. Electric sense can overcome these limitations and improve the perception of the underwater environment. Discovered in the 50’s ([7]), the electric sense, is the ability to obtain the perception of an environment based on the perturbation of an electric field. Some weakly electric fish, like the Gnathonemus petersii, are able to generate an electric field around them to navigate in turbid water, to communicate or to look for hidden food under the sand. They polarize their body relatively to an electric organ discharge that is located just before the tail (figure 1 (a)). Then, they use a large number of electro-receptors along their body to detect a deformation of the electric field and they use a brain process to create a 3-D picture of their environment ([4] and [9]). This capability to create an electric field and sense an object in its surrounding is called active electro-location (figure 1 (b)). Some others aquatic species, have the ability to detect theses electric fields but also electric potentials generated by muscle or nerve. So, through some electro-receptors they are able to detect others animals. This capability to locate electrogenic sources is called passive electro-location. In [6], A. Kalmijn presents this behavior with electro-receptive animals such as sharks (Scyliorhinus canicula) and rays (Raja clavata) which are able to detect some prey buried under the sand. D. Hopking ([5]) compares this passive electro-location as a passive listening. The electric fish and electro-receptive animals give us a new way of sensing in challenging environments. Work has been made to investigate on the application of the electric sense for a robotic purpose ([1] and [8]). Because of this sense is omnidirectional, it is particularly appropriate for the obstacle avoidance. But it can also be used for obstacle detection and localization, this behavior is called electro-location. Then, [3] present a new approached of electro-location using a reflex navigation based on passive electro-location in real fish. This technique allows to avoid any electrically contrasted object or to seek a particular contrasted object while it’s avoid others. Even if we don’t know the position or the orientation of the source of the electric field in the environment, we track the current lines generated by a exogenous electric field to guide the probe. In the following, we propose to use the passive submodality of the electric sense using a reflex navigation [3], to provide an autonmous docking for ROV wich will be also be able to avoid abstacles. This approch does not need a model of the environnement and it’s cheap to implement. It just need to equipe the docking station of an active probe which will generate the electic field. I. EXPERIMENTAL SETUP AND CONTROL We want to demonstrate the ability of a ROV equiped with the electric sens to achieve a task of docking in a complexe scene. The experiment has to show how it’s possible to guide a ROV in a autonomous way from any initial position to its docking station with or without object in the scene. The docking station is equiped with a two electrode probe which generates an electric field (figure 2 (a)). The ROV used in this experiment is a probe equiped with 6 electrodes. Those electrodes measure the electric field generated by the docking station. The ROV and the docking station are placed in a 1 *The work presented was performed as part of the strategic partnership between CEA Tech, Mines Nantes and ARMINES, funded by the region Pays de La Loire, in the frame of the Regional Technology Transfer Platform (PRTT) of CEA Tech. 1S. B. Ferrer is with the Commissariat a l’Energie Atomique et aux Energies Renouvelables (CEA Tech Pays de la Loire), Technocampus, 44340 Bouguenais, France, [email protected] 2V. Lebastard and F. Boyer are with the IRCCyN Laboratory, Mines de Nantes, 44307 Nantes, France, [email protected] [email protected]