These Incredible Spherical Robots Climb Walls, Assemble Together Like Voltron

Robots today tend to excel when they’re designed for a single purpose that is known in advance, but have trouble adapting to new or unforeseen circumstances. A major reason for that is that traditional programming techniques aren’t great for improvisation. Thankfully, artificial intelligence is progressing rapidly and is giving robots the ability to work out their own solutions to new problems. The physical structure of the robot, however, still limits its ability to implement those solutions. That’s why a team of researchers from Chinese University of Hong Kong in Shenzhen have created unique spherical robots called FreeBOTs that can assemble together like Voltron when needed.

Each FreeBOT resembles a large, undecorated Christmas ornament. These softball-sized robots have an outer ferromagnetic metal shell. Individual FreeBOTs are able to roll around like a Sphero robot, thanks to internal motors. Similar to a hamster running in a ball, a FreeBOT can easily roll across smooth surfaces. The robots are also equipped with an internal permanent magnet that helps them climb inclined surfaces if they are ferrous. The robots are light enough and have strong enough magnets that they’re even able to scale completely vertical walls. That capability alone would make FreeBOTs worthy of notice, but it’s their ability to assemble themselves into a single larger robot that makes them especially noteworthy.

Because each FreeBOT is encased in a ferromagnetic shell, many of them an easily connect together magnetically. They can form chains or multiple FreeBOTs can attach to a single central FreeBOT—or any combination of the two. Once connected, the magnetic contact point acts like a joint that offers a lot of freedom. One FreeBOT is essentially driving around on the surface of whatever FreeBOT it is currently connected to. This results in far more flexibility and versatility than you get with conventional joints. The team’s work all seems to have gone into developing the mechaniacal capabilities of these robots, and they are currently just remote controlled and don’t have any sensors for understanding their environment. In the future, they will benefit from more sophisticated autonomous control. If the team can devise a way to track the exact location of each FreeBOT in real time, the robots could work together as a swarm to accomplish complex tasks.