A View from Christopher Mims
Divers Command Robot with Waterproof Tablet
Water is an unforgiving medium for communication, but a new approach using visual tags shows potential.
A new, surprisingly simple communications method between divers and underwater robots is a significant addition to the the small arsenal of tools available to operators of robotic submersibles, or ROVs. The method could enable new kinds of on-site diver/ROV collaboration in difficult situations, such as military operations, environmental remediation and aquaculture.
Underwater communication between divers and submersible robots, or ROVs, is an almost intractable problem. Unlike control of terrestrial robots, radio waves aren’t an option because they don’t travel far under water and are easily distorted. Sonar (sound) is low bandwidth and requires enormous amounts of power (enough to damage marine life, in some cases) and communication with light is foiled by “aquatic snow” - all the particulate matter floating in the water column.
“The problem is really quite bad, to the point that many militaries don’t operate ROVs and divers in the same environment, because the communication and coordination is too difficult,” says Michael Jenkin, senior author on a paper forthcoming from York University called Swimming with Robots: Human Robot Communication at Depth.
Jenkin’s and lead author Bart Verzijlenberg’s solution to the problem is a waterproof tablet that can display the same sort of two-dimensional bar codes, or tags, already in use on some products, as well as advertisements and stickers designed to be read by smart phones.
Flashing these tags at the AQUA robot’s underwater camera allows communication that is robust and high-bandwidth relative to other untethered underwater communication modalities. In the example at left, a 6x6 tag can display 36 bits, but is redundantly displaying only 10 bits, which correspond to a command stored in the robot’s memory.
A video of the robot operating in tethered mode (below) shows its ability to react to these visual tags in real time and transmit video back to the underwater tablet. In an autonomous mode, a video feed would be impossible, but the robot would be able to autonomously carry out tasks and report back to a diver directly, rather than traveling all the way to the surface in order to communicate with its handlers.
To date, the system has been tested in the open ocean and swimming pools, but future plans include penetration of wrecked ships.
AI is here.
Own what happens next at EmTech Digital 2019.