Director of the Robotic Lab at Stanford University (California), Oussama Khatib is the inventor of the underwater archaeological robot Ocean One. He explains the improvements made to the new Ocean One K prototype for working in great depths.
What was the goal of this second Ocean One campaign?
The first campaign, which we carried out in 2016, near Toulon, demonstrated that Ocean One was able to recover and manipulate objects underwater. However, the wreckage of The moon on which we had then worked, is at a relatively shallow depth: barely 91 meters. However, for it to be of interest, our technology had to be able to be used at levels inaccessible to divers. That is to say at 1,000 meters at least. Giving the robot this capacity for immersion in the abyss was a very difficult challenge, which forced us to revise the design from top to bottom. But, finally, it works: on February 17, near Cannes, the new improved version of Ocean One K managed to transport a plate bearing an inscription up to 852 meters, at the bottom of an underwater pit. It’s great ! It’s better than a Palme d’Or: it’s a world first. No one has ever achieved such a feat.
What modifications have you made to Ocean One to adapt it to great depths?
They are considerable. We had to replace the flotation materials with “syntactic” foam, a new type of composite incorporating small hollow glass balls, manufactured by a French company. Then we tackled all the electronic and mechanical issues. From the outset, we had decided to equip Ocean One with “elastic” robotic arms or compliants, as we say in English. The latter manage to operate well below sea level, because they are filled with oil whose pressure is kept equal to that of the environment thanks to two “compensators” placed on the rear part of the robot’s body. The system worked perfectly.
“We have revised the design of the hands. Those of 2016 had three fingers, which proved to be insufficient for the manipulation of tools »
But we cannot claim to reach great depths without considering some additional precautions. From the design to the nature of the electronic components and sensors used, including the control devices, everything has been changed. Finally, we have revised the design of the hands. Those of 2016 had three fingers, which proved to be insufficient for handling the tools. At Stanford, we dedicated ourselves to adding an extra finger. And the team from the Italian Institute of Technology and the University of Pisa, with whom we collaborate, have developed a soft hand, a prototype that can be activated by a single motor of an “artificial hand” with five fingers and 21 joints, specialized in grasping objects. Depending on the task at hand, we call upon one or other of these manipulative organs.
You have 38.67% of this article left to read. The following is for subscribers only.