Titan implosion: Is artificial intelligence the future of deep-sea exploration?

In June, the submersible Titan, carrying five sightseers to see the wreck of the Titanic, exploded thousands of meters below the ocean's surface, highlighting why humans know more about the surface of other planets than the depths of Earth's oceans.

Oceans cover more than 70 percent of the Earth's surface. However, as the Titan disaster showed, this underwater world is a challenging place to explore. It's a vast space. The deepest point underwater, the Challenger Deep in the Pacific Ocean, is 11,000 meters deep, higher than Mount Everest.

Light can't penetrate so deep. Still, this little-known world is crucial to the planet's future. The ocean is closely linked to the Earth's climate, and a better understanding of the ocean could provide potential solutions to climate change. In addition, new species of plants and animals are being discovered in the deep sea.

The seabed is also home to battery metals such as cobalt, copper and manganese, which are critical to the planet's clean energy transition. Despite warnings from environmentalists about damage to fragile Marine ecosystems, the deep-sea race continues, with companies and countries looking to the seabed for resources.

The Norwegian government wants to open up an area of the seabed larger than Germany for mining. India, which in August became the first country to land a spacecraft near the moon's South Pole, announced a mission called "Samudrayaan" (Sanskrit for "sea vehicle") that aims to sail a submersible capable of carrying three people to a depth of 6,000 meters by 2026. And China is building an icebreaker with a submersible designed to reach and explore the Arctic seabed.

Is it too dangerous to explore these depths? Where does technology stand? What's next for the submersible?

Short answer: Even after the Titan disaster, underwater exploration will likely continue. However, small submersibles, often unmanned and powered by artificial intelligence, could be the future, using new technology to charge underwater and operate continuously for months or even years. But there are still some technical hurdles before we can get there.

Uncharted depths

Despite decades of scientific and technological advances that have enabled humanity to send exploration missions to distant planets, only about 25 percent of Earth's ocean floor has been mapped to date.

Still, it represents a significant shift, with only 6% of the seafloor mapped by 2017.

"So we've seen a huge acceleration in the last few years," said Jamie McMichael-Phillips, director of the Seabed 2030 project, which aims to map the entire ocean floor by 2030. Still, there is a long way to go."

Seabed 2030 does not normally map itself. It searches the archives of governments, research institutions and companies for unpublished maps of the ocean floor. Among other things, it is trying to convince other ships to use sonar systems to map the ocean floor and share data with them.

Sonar is an ancient technology, first invented in the 1910s. It uses sound waves to determine what's going on underwater and what the ocean floor looks like. With this technology, surface ships can map even the deepest parts of the ocean. The Seabed 2030 initiative turns such data into maps and makes them public.

"A range of ocean processes depend on the shape of the ocean floor, and we need this information to better understand climate change and biodiversity issues," McMichael Phillips said.

The challenge with this process is that it is slow and time consuming. A ship full of crew would need to travel the world and scan the ocean floor using sonar.

"It's a very, very slow process," McMichael Phillips said. "The game changer is going to be driverless technology, where you can operate boats almost 24 hours a day, seven days a week, without anyone on board."

Artificial Intelligence is the future

That's why Marine researchers have high hopes for AI. Ships such as submersibles operating autonomously could take up a lot of the manpower needed to explore vast oceans.

Helge Renkewitz, a researcher at the Fraunhofer Institute in Germany who studies underwater robots, said: "When you need to examine specific objects, such as the base of an offshore wind turbine, remotely operated underwater vehicles controlled by a human pilot work very well. But if you want to explore large areas of the ocean floor, autonomous submersibles are the future."

Autonomous submersibles powered by artificial intelligence will minimize the risk to human life from deep-sea exploration and be able to map the ocean floor faster. But the researchers' ideal goal is to go one step further and build submersibles that can explore indefinitely, thus speeding up the process of scanning the deepest parts of the Earth.

Renkewitz says this is difficult because there are some engineering challenges in the deep sea.

First, saltwater is corrosive, which makes it difficult for submersibles to survive intact for long periods of time, unless they are made of high-tech materials such as titanium steel. And then there's the pressure. The deeper the water, the greater the pressure on the object. This was fatal for the "Titan" submersible.

"At the depth of the Titanic, nearly 4,000 meters down, a ship was subjected to 5,689 pounds (2,580 kilograms) of pressure per square inch," Lenkewitz said. That's 400 times the average pressure we experience at sea level.

In addition, autonomous submersibles face challenges when navigating deep underwater.

On the surface, autonomous submersibles can use sensors to look around and recognize things. It can also rely on precise satellite positioning systems such as GPS. Autonomous submersibles do not have these luxury features.

Because the light from deep in the ocean is insignificant, it can only see very close to itself. Sonar can help it see further, but it can only detect objects in very specific directions. On top of that, the lack of satellite connectivity underwater makes it difficult for the submersible to find its position. Researchers use complex calculations to track the spacecraft's position, but these aren't always accurate.

"There's always an error rate with these location estimation algorithms," Lenkewitz says. The longer you spend underwater, the worse the errors. After just a few hours, you could be hundreds of meters away from where you think you are, depending on the quality of the sensor."

Continuous exploration

Another challenge for long-term submersibles is energy. These submersibles need electricity to operate, but underwater, there is no obvious power source available. Paul Kula, a professor of ocean engineering at Texas A&M University, said that solving this problem will be one of the keys to exploring the deep sea more deeply.

"Our dream is to have a permanently operational vehicle that uses renewable energy to monitor the ocean and keep us informed of any changes," he added.

Some submersibles are already taking steps to realize this vision. Underwater gliders absorb water to make them glide downward, then release water again to lift themselves up and control themselves with their wings. In this way, they can float up and down on the ocean for months. But even then, they are ultimately limited by battery life.

To overcome this, there are several options. Although the sun does not penetrate deep into the surface, autonomous submersibles can periodically surface to store energy before sinking again. But Kula said the small size of the submersible would limit the amount of solar energy it could collect.

Another scenario the researchers are considering is floating charging stations across the ocean, where submersibles can dock and recharge. What's the problem? This requires a high start-up investment.

"The initial uptick was very slow," Kula notes. You need an Elon Musk-type figure to make this happen and standardize the Marine charging connector."

Another option is to use ocean currents or hydrothermal vents on the ocean floor, although these are not always available everywhere. Kula is also developing a system that uses the thermal difference between water at different depths to generate energy. In this way, the submersible can move up and down in the water and generate the power needed to sustain itself.

Getting any such machine to work in harsh ocean conditions is not easy. Still, Kula is optimistic.

"The timing seemed right," he said. Interest and funding continue to grow, and technology continues to advance. That being said, if we funded deep-sea exploration the way we fund space, we'd already be much further along."