Exploring the Depths With Autonomous Underwater Vehicles

Rapid technological progress from virtual reality to ChatGPT brought many new innovations and promises that are widely discussed among researchers and practitioners. However, there is one area that is still largely unexplored and contains many secrets — the marine world. The oceans cover more than 70% of the earth’s surface and support an estimated 90% of the life forms of our planet. The deep ocean is considered to contain the secret origin of our lives, the oceans remain largely unknown to us due to a lack of our own capabilities, new methods, and tools. Marine robotics is one of the enabling technologies that allow us to undertake this challenging task of exploring the underwater world, but what do we know about the history and the current progress of this technology?

A Glimpse at the History of Underwater Vehicles

The history of underwater vehicles begins with Cold War. During a Cold War mission to reach the European borders of the Soviet Union, a B-52G bomber carrying four thermonuclear bombs flew over the Atlantic Ocean from North Carolina, requiring mid-air refueling over Spain. On January 17, 1966, while refueling behind the tanker KC-135, the planes collided, causing an explosion witnessed 1.6 km away. All four thermonuclear bombs were lost in the ocean. To recover them, a team of 150 qualified divers and the Cable-Controlled Underwater Vehicle were deployed, marking one of the earliest uses of remote-underwater robots for deep-sea operations. Nowadays, there are different types of underwater robots, including remotely operated and autonomous.

Remotely-controlled underwater vehicles are typically connected to a ship by a series of cables and require remote navigation, however, they are easy to deploy and require minimal maintenance. Autonomous-underwater vehicles, on the other hand, are independent of any ships and are capable of completing the missions on their own after which they can return to the original location to transmit the collected information. Even though they require much more maintenance and more resources to deploy, these underwater robots are more advanced and capable than ever before, performing tasks once thought impossible.

From Cable-Controlled to Autonomous Underwater Vehicles

One example of a currently used autonomous underwater vehicle is Argo. This is an international array of nearly 4,000 self-submerging robots. These robots drift with the ocean currents and gather data, such as temperature and salinity at different depths between the surface and mid-water level. Each float spends almost all its life below the surface transmitting its observations via a skinny black antenna to Jason, Earth’s observing satellites. Argo enables scientists to understand the ocean’s role in the Earth’s climate and predict such things as rainfall patterns, the intensity of hurricanes, and sea level rise. Some of the fleet’s floats can dive deeper than 2,000 meters, which is more than a mile down. In comparison, the world’s deepest dive on open circuit scuba stands at 332.35 meters.

2,000 meters is however not the deepest dive by an underwater vehicle. On May 31, 2009, the remotely-controlled Nereus robot reached 10,902 meters when it was sent to the bottom of the Mariana’s Trench in the western Pacific Ocean. Even though the Nereus robot was lost during the mission, it spent over 10 hours on the bottom, sending live video feedback to the ship through its fiber-optic connection with geological and biological samples collected with its manipulator’s arm. Using a fibre-optic connection instead of copper was one of the major improvements allowing ROVs (Remotely Operated Vehicles) to dive deeper and accomplish more complex tasks.

However, the emergence of autonomous robotic vehicles, including self-driving cars, drones, and domestic robots, has had an even greater impact on underwater exploration. The creation of autonomous underwater vehicles permitted them to be used for critical infrastructure protection, especially in the gas and oil industry to prevent catastrophes, search and rescue operations, surveillance, mining, data gathering, and deep water inspection. Autonomous underwater vehicles could also charge their batteries at the docking stations placed underwater, allowing for the power resources to be mostly dedicated to the missions themselves.

Nowadays both ROVs and AUVs are equipped not only with video cameras and variable lighting, but also acoustic and tracking sensors, non-destructive testing sensors, cleaning devices, and multiple single or multi-purpose work tools like bars, hooks, and even knives.

Types of Autonomous Underwater Vehicles

Echo Voyager, Boeing’s first extra-large unmanned undersea vehicle, first began operating in 2017 and already spent 10,000 hours operating at sea and transmitting hundreds of nautical miles autonomously. It is approximately the size of a school bus and can be used for oil and gas exploration and analysis of the infrastructure. Countries like Australia and the UK work on developing and deploying large autonomous vehicles for defense purposes. In addition, China recently completed the construction of the Zhu Hai Yun, an unmanned ship that utilizes artificial intelligence to navigate the sea with no crew required. So what are the types of autonomous underwater robots? First, man-portable AUVs are the smallest class that are typically shaped as torpedos and weigh around 10-50 kg, such as the Mk 18 Mod 1 Swordfish developed by US company Hydroid. They can be used for low-visibility sea explorations up to a depth of 40ft. Then slightly bigger lightweight AUVs are usually up to 227kg, such as ATLAS’s SeaWolf or General Dynamics Mission Systems’ Bluefin-12S which can carry multiple payloads and perform various mine operations. The heavyweight types of AUVs generally weigh between 5,000kg and 10,000kg and are used for longer missions that require endurance of 40-80 hours. Finally, extra-large AUVs are mostly under development, such as Boeing’s Orca which is designed to outperform Echo Voyager and spend months performing duties at sea.

Challenges and new solutions of Autonomous Underwater Vehicles

The main challenges that any development of AUVs has to face include the development of suitable components that not only requires water-tight containers and equipment but also should be able to withstand high pressures and low temperatures. Then it also has to overcome the communication losses and multi-path effects while operating with reduced bandwidth and low reliability of communication. Finally, long-range missions require that vehicles would be equipped with proper power supply systems and use this energy efficiently. One of the ways that scientists aim to overcome the power supply issue is to generate electricity using bio-inspired mechanisms. For example, scientists from the University of Bristol in the UK developed the Row-bot, an autonomous robot that feeds off an organic matter in the dirty water it swims in — the idea is that the robot should be able to operate indefinitely by scavenging its energy from the environment. Such solution would not only allow to generate energy but also help clean the ocean, which has become one of the main missions in recent years among multiple non-profit organizations and research groups. Some examples are US-based Clear Blue Sea or the French start-up IADYS which is committed to providing robotic solutions for removing the plastic and other trash pollution from water sources. Some other projects that scientists are working on include dealing with oceanic creatures. For example, researchers at Harvard University aim to use soft robots that could capture tiny and delicate oceanic creatures like jellyfish, collect cells and scan their genome, and then release them without causing harm. Researchers at the Queensland University of Technology in Australia are working on the COTSbot and RangerBot projects aiming to find venomous and invasive species, such as Crown-of-Thorns Starfish who feed on coral reefs and inject them with a fatal dose of bile salts in order to eliminate them from the reef.

From Underwater Vehicles to Robo-Mermaids

Many of the mentioned robots are specifically designed to perform particular tasks, however, the need to study the underwater world requires not only autonomous agency but also haptic feedback. No existing robotic submarine can dive with the skill and care of a human diver, thus, researchers working on OceanOne were inspired by human divers and created the first robo-mermaid: its head is equipped with stereoscopic vision and transmits to the humans exactly what the robot sees while its two fully articulated arms are equipped with haptic feedback system, meaning that the human pilot can feel what robot feels. The tail section of the robot carries batters, computers, and thrusters. What is more, the humanoid form allows the OceanOne to dive alongside actual humans and communicate to them any potential danger with hand gestures. Every aspect of this shape and design is meant to take the tasks that are dangerous for human divers, but crucial for either preventing underwater disasters or exploring the ocean.

If we go back to the 1966 Palomares B-52 crash, even though 150 qualified human divers searched the water for remaining thermonuclear bombs up to 120 feet with compressed air, 210 feet with mixed gas, and 350 feet (110 m) with hard-hat rigs, the bomb lay at a depth of 2,559 feet (780 m), which is below the limits of the human divers, but very easily achievable using humanoids and sea robots. In the future, advancements in sensing, communication, and AI technologies are likely to make AUVs increasingly capable and versatile, opening up new scientific opportunities for ocean exploration and research. With their ability to operate autonomously in challenging underwater environments, AUVs have the potential to significantly expand our knowledge of the ocean and its ecosystems, as well as help address pressing environmental and societal challenges such as climate change, marine pollution, and natural disasters.