Swarms of Robotic Bees Will Map the Deep Ocean

XPRIZE announced the semi-finalists of its Shell Ocean Discovery Prize — a competition to find the best solution to cheap, efficient, and reliable mapping of the deep seafloor — on Thursday. Twenty-two teams representing 13 countries will move on for a shot at $7 million in awards.

“What I’m really proud of the teams for is coming up with all of these different ways and approaches in which to tackle this problem,” Jyotika Virmani with XPRIZE tells Inverse. The teams must build robotic systems that can be launched from shore and autonomously navigate through the competition area. “It’s very much like a NASA setup, with mission control at the shore.”

Some teams have proposed drones that carry smaller submarines through the air and launch them out into the ocean. Others are building hybrid air/water drones that take care of both sides of the job. Some are using autonomous mini-ships from which the subs will be launched.

One surprise, says Virmani, was that more than one of the successful teams took inspiration from nature, proposing swarms of robots that will fly in a collective like insects to accomplish the task. “That was something I was not expecting,” she says. “That was really nice.”

Christopher Lewis, leader of the Eauligo team, is among them. He plans to build several hundred “marine bees” that will work as a hive to map the deep seafloor. “Just as the real bees leave their hive and go out into the fields and woods looking for flowers, and then they return and tell the other bees what they have found, our marine bee submarines do exactly the same,” he tells Inverse. “They dive down to the bottom of the ocean floor, they explore and map an area, and then they bring back the nectar, which for us is the data and images.”

The bees will operate around a mechanical “hive” — an autonomous ship that will carry the bees out to sea and launch them into the ocean. Mission control will communicate with the hive from shore, and the hive will be able to communicate with the bees — each about the size of a hand — before their dives and after they resurface. If a particular bee fails to complete its mission on a dive, the central brain will be notified and another bee can be sent to finish the job.

The idea of a swarm of simple bots has some key advantages over a single high-tech system, he says. “I think we can manage the depth better with something small and simple. I think we can get a more cost-effective solution my mass-producing a large number of simple things, rather than building one complicated thing. We have a sort of robustness as well. If one of our bees fails to complete the task, we still have the results from the other, which you wouldn’t get if you had one, large, single solution.”

“And at the end of the day, I like systems where you have very few rules, but that show amazing complexity — like real life, in fact. Like bees, like ants — individually they have very simple rules, but together they can be used for tasks where it’s extremely complex.”

For the first round of testing, set to take place this fall, the teams will each have 16 hours to map at least 38 square miles of sea floor at a depth of 6,500 feet. They will be producing bathymetric maps — like topographical maps but for the seafloor — and must achieve a resolution of no less than 16 feet horizontally and 1.6 feet vertically. Each team must also bring back at least five photos of geological, biological, or archaeological features their bots came across.

The larger goal of this enterprise is to get new technologies out into the wild that will trace the unexplored volcanoes, reefs, and treasures of the deep — including an estimated three million shipwrecks.

One of the major costs of seafloor mapping, the way it is done today, is sending ships out to sea from which to launch underwater robots. These technologies will have an edge, since they can be operated remotely. “The goal is that, after a few years, there will be a whole slew of technologies that are capable of mapping the deep sea, and that can be sent out quite cheaply,” says Virmani. “The fact that we’ll have all these different types of technology means that people who need access, or want to get down to the seafloor, or those who are interested in getting those maps out, have more options.”

Dale Wakeham, leader of Team Tao, is also putting stock in the divide-and-conquer approach. “Fire ants in the Amazon are individually unable to swim but as a colony, they can cross water,” he tells Inverse by email. “From the shore, they start walking over each other to form a raft in the water. The ants in the rear then continue to walk over this ‘raft’ and attach themselves at the front, thereby extending it. This process is repeated until the ants have bridged the water. Thus, the relatively simple ant forms part of a complex system that is able to complete an otherwise unachievable goal.”

The fire ants offer a philosophical grounding for the team’s work, although they are not trying to mimic the specific behavior of the insects, he says. Rather than using conventional components to build a single high-tech robot, Wakeham’s team plans to build a swarm of low-tech bots that gain their intelligence in their ability to work as a collective. Like the ants, you can afford to lose a few individuals without affecting the operations of the group. The system is therefore easy to scale up and also less vulnerable to the sort of technical difficulties that might leave a single underwater drone beach-bound.

After the first round of testing, a maximum of ten finalists will move on to a final round of testing at a crushing depth of 13,000 feet. A bonus prize, sponsored by NOAA, will award teams building systems that can sniff out a chemical signature underwater and trace it to its source. The award, though, is just the beginning. The ultimate prize will be the ability to push further in our understanding of Earth’s final frontier.