The ocean floor, the last frontier of earth, will reveal itself thanks to swarms of robots talking to one another with acoustic pulses. Or so Delft researchers believe.

Sunken drums on the seafloor leaking chemicals, shipwrecks and containers dropped from a ship. A team of Portuguese scientists – or more specifically, their robots - had little difficulty locating these objects during recent experiments in the harbour of Porto. The robots formed an effective search team thanks amongst others to algorithms developed by Dr. Hamid Ramezani, who defended his PhD research about underwater acoustic localization this summer at the EWI faculty.

Ramezani and his supervisor, Professor Geert Leus, took part in Project Noptilus, led by the Centre for Research and Technology Thessaloniki (Greece). Noptilus is an acronym for autoNomous, self-learning, OPTImal and complete Underwater Systems.

Exploration of the deep sea nowadays involves submarines with human pilots or robots that are either attached to a boat or need to surface regularly to transmit their discoveries. That makes the exploration a tedious task. And often unsuccessful. The ocean floor search for airplane MH370 has been unsuccessful for months.

Torpedo shaped robots

There is a more efficient way, according to Ramezani. It involves autonomous robots. The three torpedo-shaped robots, or autonomous vehicles, which the Noptilus-researchers deployed, communicated with one another, and with a fixed station in the harbour, using acoustic pulses. Kind of like the sonar used by submarines.

"The ultimate goal of Noptilus is to unleash swarms of robots that can communicate with one another and form vast underwater networks", explained Ramezani. "Radio signal frequencies, used for instance in Wi-Fi, fade out in the water within a couple of meters. We, therefore, have to fall back on acoustics."

Ramezani, and his colleagues within Noptilus, are not the only ones with this vision. Many more groups are working on techniques to disclose the oceans with swarms of robots.

Scientists from Italy, Germany, Portugal, The Netherlands (University of Twente), Turkey and the United States, for example, teamed up in Project Sunrise (Sensing, monitoring and actuating on the UNderwater world through a federated Research InfraStructure Extending the Future Internet). Their slogan: Building the Internet of Underwater Things.

Another example is the Austrian/ Belgian Subcultron project involving "mussel" robots that sit on the sea floor and act as a coordinating grid.

The "mussels" float to the surface when they need to talk to a base station. They chat to "fish" explorer robots that can also talk among themselves, reducing the distance any signal need travel. The result is a "dynamic seabed-exploring carpet, which can then slowly crawl through a larger habitat", project leader Thomas Schmickl of the University of Graz told New Scientist.

Salinity, pressure and temperature

The challenges are vast. The propagation speed of sound in the water depends on salinity, pressure and temperature. If you assume the speed to be constant the robots will not be able to determine their position accurately. Part of the Delft research consisted of developing algorithms that take these variables into account.

Another problem is encoding the information. "With radio, you can send in the order of megabits per second of information. With acoustic modems the data transmission rate would be around hundreds to a only a few kilobits per second, a factor of a thousand less", said Ramezani. "And you can't broadcast the information with too much power either because you might interfere with the echolocation of whales and dolphins, depending on your transmission frequency."

That is why an extensive network is needed. Robots could send information at low amplitude through to each other like a relay baton. Via a series of nodes, the information can get to central data points.

"Fibre optics are already crossing the oceans. Why not use this infrastructure to make fixed nodes on the seafloor that can communicate with the robots. Or the robots could communicate with buoys at the surface that are in contact with satellites", said Ramezani.

And then there is the challenge of the energy supply. Ideally, the robots would be able to harvest energy somehow from the oceans. The Subcultron robots get their energy by docking with floating "lilypads" of solar panels, but at great depths, another solution is needed. Schmickl told New Scientist that his robots might one day use fuel cells powered by bacteria.

"Sustaining the tremendously high pressure at the deep ocean bottom is another issue we have to tackle," said Ramezani. "It is difficult, but do really think this is where we are heading towards; an underwater internet comprised of robots."

Outer space

Moreover, the deep sea isn‘t the only difficult-to-reach region the research group wants to access. There’s outer space too. Radio astronomers dream of deploying swarms of nanosatellites in outer-space that detect radiation from distant cosmic sources. Enabling these satellites to form a network requires though mathematics. This is the work of Raj Thilak Rajan, who will be defending his thesis Relative Space-Time Kinematics of an Anchorless Network later this year. He too collaborates with Prof. Geert Leus.


This article was changed on August 18th 2016. At first it stated that Project Noptilus, was led by the University of Porto. However, only the demonstration in the harbour of Porto was led by the University of Porto. The actual project is led by the Centre for Research and Technology Thessaloniki (Greece).