Georgia Tech researchers teach robots how to follow the leader and work together.
We can all imagine the days when manmade robots equipped with limitless knowledge and abilities will somehow become beings that can think on their own—thus, leading to our impending doom. People have managed to develop robots that can walk upright, respond to human signals, and even do marvelous things like ride a bike, but we have yet to encounter a sentient being that’s completely made of metal and silicon (at least not first hand).
Skipping the development of a single robotic entity that can function on its own, researchers at the Georgia Robotics and InTelligent Systems Laboratory went straight to developing a unit of bots that can function by following the instructions of its leader.
In this systematic robo-unit, the leader is the one that gets the initial set of instructions, which it then relays to neighboring bots (the followers). Conditions such as distances between the bots will determine which of the commands the drones will execute sent via the leader. This sounds simple enough, but the complexity in the whole system lies in the algorithm that will determine which bot will do what.
“The challenge…lies in building topologies tailored for applications that specify the total number of such groups, as well as the cardinality of each group. We’re currently working on developing low complexity algorithms for building such networks,” said the researchers.
To demonstrate a simple use of the system, the researchers set up the robots to play an extremely short rendition of Beethoven’s “Fur Elise” on a piano table. The tune probably won’t win the bots an Grammy or VMA any time soon, but the prospect of robots working together in unison and good efficiency is evident.