With all the buzz about robot aircraft and robot cars, we should not forget that the robots are coming on the sea, as well as the land, and the air.
In fact, sea-going robots make a lot of sense for many reasons. The ocean is dangerous (and big!). But ships move slowly (compared to air craft) and it’s not too crowded (unlike city streets).
A couple of recent reports show what might be done.
Jeremy Hsu reports on US Navy trials with swarms of robot boats (“Swarmboats”), operating as a team. Directed by and reporting to human sentinels, the robots spot, identify, and close with intruders. This is cost effective (you can buy a lot of swarmboats for the cost of an aircraft carrier!), extends the reach of the fragile and hard to get humans, and creates a resilient, distributed asset.
The underlying control system and software is designed to be installed on pretty much any watercraft, which is a really good idea. This is obviously cost effective (building custom boats is impossibly expensive), and in any case the design of boats and ships doesn’t need to be “disrupted”. We have been building them for millennia.
(I am reminded of the time I remarked to an engineer from a major agricultural machinery company, “the future is lot’s of small robots, reconfigurable to different tasks”. The advice was not welcome by a manufacturer of really big equipment.)
Another project of interest developed a humanoid robot for underwater salvage. Technically, this is a remote operated vehicle (ROV), not fully autonomous. But ROVs are useful to the degree that they offer “high level” capabilities to the operator. I.e., the ROV needs to know how to move and do stuff, so the human can concentrate on what to do and what is important.
Oussama Khatib and colleagues reports on the Ocean One is a humanoid submarine ROV . There are plenty of underwater ROV’s, of course, but most look like refrigerators with arms. Ocean One dose similar things, but has a humanoid body. The control system is an immersive display, with a haptic harness that ties the head and arms of the operator to the head and arms of the ROV. Thus, it is and “avatar”, closely matched to the human operators body.
“The key ingredients to its success were: 1) the whole-body control and interaction skills of the robot balanced with the expertise of the human and 2) the haptic-visual interface bringing a heightened sense of presence.” (, p. 28)
It certainly looks cool, though I’m pretty sure that operators could develop similar sense of “presence” with any ROV that had the same interface, regardless of whether it looked humanioid. A refrigerator or a sphere would work, provided that the “head” and “arms” map to the human movement.
Nevertheless, it makes sense that a ROV that is intended to operate as an avatar should look like and avatar. This will become even more useful when operating in groups and with divers. How do I know that that 7 DOF hunk of machinery is operate by George? Well, it should look like George, no?
- Evan Ackerman and Erico Guizzo, How Stanford Built a Humanoid Submarine Robot to Explore a 17th-Century Shipwreck, in IEEE Spectrum – Automation. 2016. http://spectrum.ieee.org/automaton/robotics/humanoids/stanford-humanoid-submarine-robot
- Jeremy Hsu, U.S. Navy’s Drone Boat Swarm Practices Harbor Defense, in IEEE Spectrum – Automation. 2016. http://spectrum.ieee.org/automaton/robotics/military-robots/navy-drone-boat-swarm-practices-harbor-defense/
- O. Khatib, X. Yeh, G. Brantner, B. Soe, B. Kim, S. Ganguly, H. Stuart, S. Wang, M. Cutkosky, A. Edsinger, P. Mullins, M. Barham, C. R. Voolstra, K. N. Salama, M. L’ Hour, and V. Creuze, Ocean One: A Robotic Avatar for Oceanic Discovery. IEEE Robotics & Automation Magazine, 23 (4):20-29, 2016.