Tracking Sharks With Robots

Scientists have been tracking sharks using robots for years. However, a new design allows them to do this while following the animal. Biologists from Mote Marine Laboratory and engineers at Harvey Mudd College developed the system using off-the-shelf components.

It is able to withstand a pull-off force that is 340 times stronger than its own weight. It is also able to detect and adjust its pathway based on changing objects in the home.

Autonomous Underwater Vehicles (AUVs)

Autonomous underwater vehicles (AUVs) are robotic machines that, according to their design they can drift, move or glide through the ocean without any real-time guidance from human operators. They are equipped with a variety of sensors to record water parameters, and to explore and map ocean geological features, sea floor habitats and communities, and more.

They are controlled by a surface ship with Wi-Fi or acoustic connections for sending data back to the operator. AUVS can be used to collect spatial or temporal data and can be deployed as a large team to cover a larger area faster than one vehicle.

AUVs are able to use GPS and a Global Navigation Satellite System to determine their position around the globe and the distance they've traveled from their starting location. This positioning information, along with sensors for the environment that transmit information to the onboard computers, allows AUVs to travel on a planned route without losing sight of their goals.

Once a research project is complete, the AUV will be able to float to the surface and then be recovered on the research vessel from which it was launched. Alternatively, a resident AUV can remain underwater and conduct regular, pre-programmed checks for months at a time. In either scenario an AUV will periodically surface to transmit its location via the GPS or acoustic signal, which is transmitted to the surface vessel.

Certain AUVs are able to communicate with their operators on a continuous basis through a satellite connection on the research vessel. Scientists are able to continue their research on the ship while the AUV collects data underwater. Other AUVs may communicate with their operators only at specific times, for instance, when they need to refuel or to verify the status of their sensors.

Free Think claims that AUVs are not only used to collect oceanographic data but can also be used for the search of underwater resources, like gas and minerals. They can also be used as part of an environmental disaster response plan to assist with search and rescue operations following tsunamis or oil spills. They can also be used to monitor subsurface volcanic activity and to monitor the health of marine life, such as coral reefs and whale populations.

Curious Robots

Contrary to traditional underwater robots, which are programmed to search for one specific feature on the ocean floor, these curious underwater robots are designed so that they can scan the ocean floor and adapt to changing circumstances. This is important because the conditions beneath the waves can be unpredictable. For example, if the water suddenly gets warmer, it could change the behavior of marine creatures or cause an oil spill. Robots that are curious are designed to quickly and effectively detect changes in the environment.

One team of researchers is developing an innovative robotic system that makes use of reinforcement learning to teach the robot to be curious about its surroundings. The robot, which looks like a child wearing yellow clothing and a green arm is able to recognize patterns that might signal an interesting discovery. It also can decide what it should do next, in relation to the results of its previous actions. The findings of the study could be used to create a robot that is capable of learning and adapting itself to the changing environment.

Other researchers are using robotics with a curious nature to study areas of the ocean that are too risky for human divers. For example, Woods Hole Oceanographic Institution (WHOI) has a curious robot named WARP-AUV that is used to search for and research shipwrecks. This robot is able recognize reef creatures and discern fish and semi-transparent jellyfish from their dim backgrounds.

It takes a long time to learn to perform this. The brain of the WARP-AUV has been trained by feeding it thousands of images of marine life, so it is able to identify familiar species on its first dive. In addition to its abilities as a marine detective, the WARP-AUV has the ability to send topside supervisors real-time images of underwater scenes and sea creatures.

Other teams are developing robots that learn with the same curiosity humans have. A team at the University of Washington’s Paul G. Allen school of Computer Science & Engineering, for instance, is examining how to teach robots curiosity about their surroundings. This group is part of a three-year program by Honda Research Institute USA to develop machines that are curious.