Acoustic telemetry can provide important knowledge on marine animals daily movement patterns and behavior, migration patterns in marine protected areas, comparative behavior of wild and cultured fish, habitat use, and data for defining essential fish habitat. These studies on marine animal behavior are instrumental to validate and solve a number of challenging problems in marine biology, as they complement other marine telemetry methods such as pop-up satellite archival tags. Decreasing size of transmitters along with increasing reliability of the equipment resulted in applications with smaller and smaller species, in salt or brackish water.
In underwater robotics a number of techniques resorting to the emssion of acoustic waves have also been used to track remotely operated vehicles, autonomous underwater vehicles, and unmanned underwater vehicles. In mission scenarios where global positioning systems are not available, Ultra-short baseline positioning (USBL) systems aiding Inertial Navigation Systems (INS) have been successfully developed and used. The USBL is composed of an array of hydrophones with a pre-specified structure and the INS is based on measurements from triads of low cost calibrated accelerometers, rate-gyroscopes, and magnetometers, complemented with data from depth cells, Doppler velocity logs, and fiber optical gyros.
The goal of this project is to endow the scientific community with new moderate cost robotic tools able to track multiple tagged marine animals supported on USBL aided INS systems. These advances are feasible according to results of a recently completed FCT project RUMOS – PDCT/MAR/55609/2004. Resorting to the low rate (1/30 Hz to 1/90 Hz) acoustic data emitted from commercially available tags and from high data rate (100 Hz) measurements from inertial MEMS sensors, installed onboard the robotic tools, (sub-) optimal nonlinear estimation techniques will be able to provide estimates on the trajectories of the multiple targets at rates of 1Hz to 1/5 Hz, with metric accuracy during the sea missions and sub-metric accuracy after post-processing.
The Robotic Tools will be designed mainly by the IST team and will be validated in sea missions with augmenting degree of risk and complexity, involving several resources and strategies defined by the CCMAR team, in close collaboration with the partners from HMS-Stanford and DPAq-UFRPe, namely:
1) Installation of the SURFACE ROBOTIC TOOL (SRT) including an USBL aided INS system, energy sources, GPS receiver, and UHF communication modems, able to monitor an area. Examples of missions: i) detection and identification of an acoustic tag, moored at the sea bottom, ii) range and accuracy tests (e.g mooring a tag and move the SRT, with the help of a ship), iii) sound wave velocity independence test, resorting a pre-calibrated moored tag, iv) mission with several tags, carried by human SCUBA divers in a pre-specified mission, and v) presence of one or more tagged marine animals (captive and free).
2) Simulated situations where the scientists receive reports from the SRT on the number of animals present, the identification codes, telemetry data, and full 3D trajectories. Studies on detection success (false negatives/missing rates), distance and sea disturbances impact. Alarm and warning generation tests from the SRT to the shore station.
3) Use of the PORTABLE UNDERWATER TOOL to provide an estimate on the target trajectory, helping the SCUBA diver to find tagged marine animals and equipment. The tool is also composed of a USBL aided INS system, with the capability of emitting a specific signal that allows the surface craft to track the diving operation. Preliminary tests with one fixed tag in the bottom, with a human SCUBA diver carrying the tag in a pre-specified trajectory, and with one or more marine animals will be conducted.
The development of these robotic tools allows for the generalization to marine equipment and SCUBA divers of the idea of “black boxes”, commonly used in aircrafts, thus avoiding the common loss of expensive and rare devices.
Data from the robotic tools can be post-processed resulting in full 3D position and velocity trajectories with data rates and accuracies not possible from the commercially available systems.
