Zhi Dongxi (public number: zhidxcom) edited by | Wang Ying
Introduction: The new underwater robot can transform into a human form in 30 seconds and complete deep-sea tasks without manual control.
A few months ago, Evan Ackerman, a special editor of IEEE Spectrum, came to NASA's pool for training astronauts to see the Aquanaut, a robot submarine built by Houston Mechatronics, which can skillfully transform into a human form in the water.
Aquanaut's exterior is bright orange and shaped like a miniature submarine. On the outside, it looks no different from other driverless underwater robots, but it's equipped with sensors for cellphone data and thrusters for forwarding.
Aquanaut represents an entirely new design, and its creator, Houston Mechatronics, hopes that Aquanaut will revolutionize undersea robotics.
Aquanaut's transformation into a humanoid is like the scene in the movie Transformers, with the top of the robot's shell rising, two huge arms unfolding from either side, a wedge-shaped head filled with sensors spinning into place, and within seconds, the transformation is complete. By this time, the originally sleek submarine had turned into a half-humanoid robot, ready to start working.
Evan Ackerman entered the pool and felt floating in weightlessness, with only mission control communicating with him through headphones. Not far from him, two astronauts were practicing spacewalks, but he didn't have the energy to visit the space training process, and the main purpose of his trip was to see the underwater robot Aquanaut.
Traditional unmanned underwater robots are usually divided into two categories, torpedo-type free-swimming submersibles for remote survey tasks, and quad square remote control machines for supporting ships and for underwater maneuvering. Houston Mechatronics hopes to combine these two models into a single robot, an unprecedented bold attempt.
Aquanaut is designed primarily to provide repair services for subsea oil and gas installations that require companies that own and operate these facilities to spend significant amounts of money to inspect and maintain them. In the extreme working environments of the seafloor, the robotics they rely on have not fundamentally changed in decades.
More than 24 of Houston Mechatronics' 75 employees have worked for NASA, and bursting creativity in extreme environments is what they do. Nic Radford, co-founder and chief technology officer at Houston Mechatronics, worked on advanced robotics projects at NASA's Johnson Space Center in Houston for 14 years. "It's more difficult to go underwater than it is to go into space, space is a pristine environment, and things underwater are very dynamic, and I can't conclude whether robots work underwater 10 times or 50 times harder than in space," he said. ”
Aquanaut can transform from a flexible submarine designed for long-range cruising into a half-humanoid robot capable of performing complex operational tasks. Here's how the robot changes:
1. Aquanaut travels to the undersea workplace in a simplified submarine mode.
2. Once the robot arrives at the scene, the top of its hull rises, exposing two huge arms and a wedge-shaped head.
3. The head carries a stereo camera, 3D sensor and sonar system, rotating into place.
4. The robot unfolds its powerful arms and is equipped with force sensors and claw clamps.
Since the founding of Houston Mechatronics in 2014, Nic Radford and other co-founders Matt Ondler and Reg Berka have raised more than $23 million in venture capital.
Most of the work we see and hear about the offshore oil and gas industry is done on platforms where people conduct underwater drilling from the surface of the sea. The platform is the most obvious part of the whole process, but there is also a lot of complex infrastructure on the seabed.
Nic Radford said: "Putting robots in remote places and letting them do useful work in a harsh data environment is best suited to this big problem: working offshore." ”
The wellhead of the seabed is covered by metal components that control the flow of hydrocarbons to the surface. These structures are covered with pipes, valves, manifolds, and instruments, some as tall as four stories, and are often referred to as "Christmas trees."
In order to perform routine maintenance on the wellhead, or to change the output of the well, some valves on the "tree" must be turned, and in deep-water wells below 300 meters, divers are usually unable to operate, and the only way is to use a robotic submersible.
For decades, the established procedure for working on deepwater wells has been to send remotely operated underwater vehicles (ROVs) to well sites. However, in addition to sending the ROV itself, it is also necessary to dispatch a large auxiliary vessel carrying well-trained personnel as an operating base for the ROV, which has little or no autonomy and requires power and control through the operation of surface personnel, a job that is very expensive, costing tens of thousands to hundreds of thousands of dollars a day.
Houston Mechatronics' plan is to reduce the need for surface control by underwater robots, and Aquanaut does not require cables or support boats. It will travel in submarine mode to its deep-sea destination, where it will transform into a human form, spreading its arms to work. Each arm of the Aquanaut is equipped with a powerful moment sensor and has eight axes of motion, similar to a human arm.
The robotic arm on the Aquanaut also has valves capable of turning the underwater "Christmas tree" and can even operate the specialized maintenance tools that the robot carries in the internal payload compartment.
Aquanaut will perform tasks that are supervised by a human operator but not directly controlled. When the task is complete, the bot will automatically return. Nic Radford said this approach will allow Aquanaut to be deployed faster and have lower operating costs than ROVs are today. He estimates that the cost could be well below half the market price of a traditional business.
After countless iterations of design, Aquanaut was finally born, but before it could actually work on the seabed, it needed to prove itself under more controlled conditions, which meant that Aquanaut also had to practice "swimming" in NASA's pool first.
NASA's Neutral Buoyancy Laboratory (NBL) can hold 23.5 million liters of water with a maximum depth of 12 meters, enough to accommodate most of the full-scale models of the International Space Station. In late March 2019, Houston Mechatronics had partially taken over the NBL to test Aquanaut.
At 10 meters underwater, Evan Ackerman carries two nitrogen tanks on his back and tracks the robot in the water. So far, Aquanaut has been successfully tested for 8 days, the only problem is that the arm has a communication failure, but Houston Mechatronics is not worried, they know that there is still a lot of work to be done to make Aquanaut up to the real working standards.
According to chuck Richards, a pioneer in subsea technology, low oil prices over the past few years have cut profits, led to increased competition among oil companies and pushed for the adoption of new technologies. Chuck Richards' Houston-based company, CA Richards & Associates, supplies equipment to dozens of subsea companies, including Houston Mechatronics.
The main advantage of Aquanaut over traditional ROVs is its unrestricted operation. Houston Mechatronics had to solve several key issues to achieve its functionality.
The first is to have the robots reach the offshore construction site without large support vessels. Aquanaut can travel more than 200 kilometers in submarine mode, and will automatically convert to ROV mode when it reaches its destination, with additional thrusters hidden inside the hull making it more maneuverable.
The process of robot deformation is another major challenge and a constant point of debate within Houston Mechatronics. Sandeep Yayathi, Aquanaut's principal engineer, believed that the benefits of building Aquanaut far outweighed the difficulties posed by its complexity, and in the end they decided to break through the difficulties to complete Aquanaut's design.
To enable Aquanaut to change its shape so drastically, the robot is equipped with four custom linear drives that separate the robot's upper and lower bodies. The additional motor, also highly customized, is mounted in a waterproof enclosure that drives the arms and head. In terms of power, Aquanaut uses a lithium-ion battery similar to that used in electric vehicles. The complete conversion from submarine to robot takes only 30 seconds to complete.
Perhaps none of these challenges are as important as designing Aquanaut's control system. Traditional ROVs have multiple real-time cameras for video transmission, and human operators can manipulate these robots in real time.
Aquanaut's only method of communication is through an acoustic modulation regulator, a proven technology that, although tens of kilometers underwater, has the disadvantages of high latency and bottom bandwidth, and its fastest transmission speed is only a few kilobytes per second. The Houston Mechatronics plan to rely on small unmanned surface ships as relay stations between robots and communications satellites to facilitate control of Aquanaut anywhere.
Houston Mechatronics plans to conduct high-level supervisory control over Aquanaut, delegating most of the low-level decisions to the robot's onboard computer. These computers run the Robot Operating System (ROS), a popular software platform for researching and commercial robots. The robot uses sensor components from the head, including stereo cameras, structured light sensors, and sonar systems, to create detailed 3D renderings of its surroundings. Aquanaut transmits highly compressed subsections to the operator, matching them to existing structural models.
The operator then issues simple commands such as "Rotate the valve 90 degrees clockwise at these coordinates." "The robot will autonomously decide how to grasp the valve and how much force to apply as it turns, and send back confirmation messages after completing the task. The operator is still directing the robot's movements, but there is no need to manually operate the robot or high-bandwidth real-time video transmission.
Houston Mechatronics' long-term plan is to sell Aquanaut's capabilities as a service, using a fleet of small robots scattered across the North Sea or california coast, where oil and gas companies simply ask for a specified task, and Houston Mechatronics arranges for the nearest robot to handle it.
Nic Radford said: "It takes about 7 people to operate a single traditional ROV, and we thought we could do the opposite, one person can operate 7 Aquanaut"
Matthew A. Franchek, a professor of mechanical engineering at the University of Houston and director of international subsea engineering research, said the risk of problems may be higher because of low-bandwidth connections and operators operating only in the middle of a cycle. "Aquanaut has a strong uncertainty and I'm concerned about failures during operations, which could have serious financial and environmental consequences," he said. While the technology is exciting, they need to prove that it works. ”
The current version of Aquanaut is primarily a demo and test platform designed for relatively shallow water with a maximum operating depth of 300 meters. While this version can be commercially operated in many parts of the world, Houston Mechatronics is already designing an enlarged version that can travel hundreds of kilometers and reach a depth of 3,000 meters that could serve regions such as the Gulf of Mexico.
Commercial operations aren't the only exploration Houston Mechatronics has done for Aquanaut. In late 2018, the U.S. Defense Advanced Research Projects Agency announced a program called Angeler to "develop a submarine autonomous system that can navigate and physically manipulate seafloor objects." The Defense Advanced Research Projects Agency, which issued the statement, attached an image of a streamlined robotic submarine with two arms, a concept that is a good opportunity for Houston Mechatronics.
Aquanaut is ready for the next NBL test. Its first open water demonstration is likely to take place during a naval technology exercise in Rhode Island in August. Nic Radford says it's his interest to engage in bold and innovative work that will prove to have a better, more cost-effective way to get underwater work done.
Original from: IEEE Spectrum
Thanks for reading. Click to follow on board and take you to the forefront of technology ~