Ivan Masmitja: "Underwater robotics allows us to bring a scientific laboratory to the site where the action happens"

Ivan Masmitja landed at the Institut de Ciències del Mar (ICM) in Barcelona a few months ago to continue deepening his research on deep-sea species, which he carries out thanks to underwater robotics and the use of techniques related to artificial intelligence. Ivan comes from the Universitat Politècnica de Catalunya (UPC), where he studied and discovered these disciplines that allow us to see what happens underwater at depths that would be impossible to access in any other way. In other words, underwater robotics allows us to shed light on the darkness of the ocean, which extends below a few hundred meters deep, where sunlight has no place. During his stay at the ICM, where he has arrived thanks to a Marie Curie fellowship, Ivan will study the use of new algorithms to train robots capable of finding the best trajectories to locate and follow electronically tagged species.

1. Why did you decide to come to the ICM?

I started collaborating with the ICM a long time ago. In fact, one of my PhD co-directors, researcher Jacopo Aguzzi, works here, so it was a natural process. In other countries, this relationship between disciplines is much more widespread. This was one of the things that surprised me most about my first stay at the Monterey Bay Aquarium Research Institute (MBARI) in California, in 2016. There, engineers, programmers, biologists or chemists, among others, work side by side under the same building. In our way, we are doing it here too, especially between the UPC (where I have carried out all my studies) and the ICM thanks to the associated unit Tecnoterra and the efforts of people like Joan Baptista Company and Jacopo Aguzzi (from ICM) or Joaquín del Río (UPC). I believe that this relationship between different disciplines is vital to advance in the field of research, so I decided to come to do a postdoc at the ICM.

2. Where did your interest in underwater robotics come from?

I was passionate about electronics when I was a child. When I was 10-11 years old, I started to design my first circuits at home and, when the time came, I decided to study electronic engineering. At that time (2007), at the UPC I could choose between the industrial branch in Barcelona or the telecommunications branch in Vilanova i la Geltrú, where I ended up going. There I met the SARTI research group and did my final thesis with Professor Spartacus Gomariz on the development of an autonomous underwater vehicle. At that moment my professional life made a 180-degree turn, and since then I have dedicated myself to technology and underwater robotics.

3. What do robots bring to oceanographic research?

Underwater robotics has become one of the most important tools for oceanography. It allows us to bring a scientific laboratory to the site where the action is taking place. It is true that satellites are also fundamental to the study of the ocean, but many times the data obtained must be adjusted with data measured in situ, and this is where autonomous vehicles appear. These robots also allow measurements to be made at lower spatial resolution and therefore make it possible to study phenomena that occur in a smaller space. Finally, they allow us to go to places where we would not otherwise be able to go, such as the deep ocean. In fact, there are robots that can go down to depths of more than 4000 meters.

4. What is the greatest difficulty in studying the ocean?

The ocean is vast, dark, deep and inhospitable. Anything that is down there is subject to great pressures, corrosion, etc. Also, electronics and water don't get along too well with each other, which means that everything has to be very well thought out and sized. On top of that, electromagnetic waves do not propagate well in water, so all the technology we have developed on land that makes our lives much easier, such as wireless connections like 4G, Wi-Fi, satellite communications or GPS positioning, is useless.

5. How can we communicate underwater?

The only way we have today to communicate underwater at long distances is by means of acoustic waves, as whales or dolphins do. However, this has its limitations, as we cannot send information much further than 4 kilometres, and the bandwidth is very narrow. This means that, for example, we cannot send video in real time. And since there is no GPS, it is very difficult to know with great precision where the underwater vehicles are, the observation platforms we are moving or the animals we are tagging. So, technologically, there are still many challenges ahead.

6. How long have there been underwater robots?

Underwater robotics was born, like everything else, from the need to exploit natural resources for our benefit. Here in Catalonia, we were pioneers in this, and in 1859 Narcís Monturiol built one of the first manned submarines, the Ictineu, which was used to collect red coral, a highly prized species at the time, off the Catalan coast. Subsequently, underwater robotics made a great qualitative leap and began to be used for the offshore oil and gas industry. At that time, ROVs (remotely operated vehicles) appeared, which allowed monitoring and manipulation tasks to be carried out at great depths much more safely than by divers and reaching much greater depths. Much later, in the 1990s, people like David Packard (founder of MBARI) began to use these tools for the study and conservation of the ocean. Today, thanks to technological developments, underwater robotics has spread around the world and a multitude of variants have appeared, such as high-performance autonomous vehicles that can spend months at sea collecting data.

7. How can artificial intelligence contribute to underwater robotics studies?

Artificial intelligence, and more specifically machine-learning, is a field that has taken off in the last decade. We have more and more machines that can process more operations per second. Moreover, today we can "train" an algorithm to "learn" to classify images or pilot vehicles. These techniques have already begun to be used to locate and track species autonomously, without the intervention of an expert. In fact, all this is set to increase, and robotics will become increasingly autonomous and capable of adapting to the environment. I am convinced that the use of multiple robots capable of coordinating among themselves to fulfill a mission will be common in the near future, and humans will have to intervene only occasionally to perform supervisory tasks.

8. What projects are you working on right now?

We are currently working on two important projects of the Spanish Ministry, BITER and PLOME. Each one has its peculiarities, but both have multidisciplinarity as a central element to advance in the study of species living in the deep sea. Within the framework of these initiatives, we are developing new technologies and methods to study the movement of species that are economically key to our territory, such as Norway lobster, conger eels or octopus. We are also studying how the use of marine protected areas -subject to special regulation- can help the recovery of their populations.

9. How can artificial intelligence benefit other disciplines?

The use of new machine-learning techniques and the improvement in underwater robotics and technology can help, and in fact are already helping, in many fields. For example, they are being used to study species at risk of overexploitation or to assess the impact of climate change on the marine ecosystem. In addition, machine-learning can be used to improve methods for monitoring and observing the effects of underwater mining or the installation of floating wind farms.

10. Any crazy idea that you would like to make real someday in your field of research?

I would like to one day have a fleet of autonomous vehicles capable of locating, tracking and studying the whales that pass in front of the Catalan coast. And why not, one day I would like to see a robot exploring the oceans beyond our planet, for example under the ice of the moons of Jupiter or Saturn.