Conceptually, the project is somewhat similar to existing warehouse robots, which use optical navigation systems that recognize markings on floors and walls. Except this research concept is a bit smellier. The Navy wants its defense-industry partners to “identify [a] chemical capable of meeting environmental and health requirements” which can act as a pheromone. Next, the Navy needs a system that can encode the chemical with data, and a system for decoding it. Eventually, the plan is to “fully develop and test modules, for a leader and follower robots, capable of operating for duration of one complete week.”
The description of “leader” and “follower” robots comes down to the Navy’s requirement that the machines are at least semiautonomous, with a human controller in charge. These robots will be tasked with carrying 1,000-pound bombs inside tight spaces, after all. Per the Navy solicitation, the leader robot will be controlled by a human who guides it along and dispenses the chemical pheromone, with follower robots picking it up, analyzing it and following along like army ants. That should help take some of the load off the aviation ordnance crews known as Red Shirts.
It’s unclear how the chemical pheromone will work beyond that. But a rough analogue came from a group of Swiss researchers who tested aerial drones with “virtual” pheromones in 2010. For the Navy, the fully-autonomous follower robots will have to detect the chemical, use data encoded within it to identify the human-controlled robot as the leader, while detecting messages along at least three “channels”: including swarm formation; speed; and direction.
It’s not going to be that simple, though. If the project works, the sniffer-robots will begin deep below the carrier’s water-line, hauling bombs from nine levels underneath the flight deck into a series of elevators, before ending up at an assembly point on the deck called the “bomb farm.” Once there, the chemicals will have to withstand winds whipping over the deck, and “must be stable enough during direct contact with petroleum products,” withstand temperatures above 200 degrees Fahrenheit, and fade after a mere 20 minutes — thereby preventing other robot swarms with different instructions from getting confused when moving down the same hallway.
But the idea of partially automating more and more of what a carrier does isn’t unique to moving bombs. By the end of the decade, perhaps, robots will be hauling bombs aboard the Ford class of supercarriers, the successor to the Nimitz class supercarriers which dominated the world’s oceans during the Cold War — and still do. The first ship of the new class, the U.S.S. Gerald R. Ford, is expected to set sail in 2015 and is being built with many of her systems automated — in comparison to the old Nimitz carriers. That’s reportedly going to reduce the crew by more than 1,200 sailors. Currently, about 70-80 members of a carrier crew are Red Shirts, of which three to six are needed to move a single bomb pallet.
But the Navy also promises the Ford class will reduce carrier costs, though the new class is a billion dollars over budget. And the Navy is using the same logic for robot swarms, claiming they will be an “elegant alternative to more costly sensors and obstacle avoidance algorithms.” That smells pretty sweet.