Swarms of robots aren’t just in movies like the 2003 Matrix Reloaded or on the pages of sci-fi novels. NYU Abu Dhabi’s Visiting Professor of Computer Science, Eliseo Ferrante, brings them to life without the movie violence and drama.
The swarm robotics concept comes from nature: studying the traits of collective organisms, how they behave, and what they can do only as a group.
“The whole idea is that at the collective level, there is a trait that an individual doesn’t have. A single ant cannot build a bridge,” Ferrante said. “A colony of ants can. A school of fish can confuse a predator by assuming a certain shape. A single one can’t do that.”
According to Ferrante, swarm robotics has two goals: to understand the science behind this naturally occurring behavior and, as he puts it, “steal” from science to design new engineering systems.
“If we can learn how these animals work, we can design an artificial system that either imitates what nature is doing or does different things using the same principles of self-organization,” Ferrante said. “This differs from traditional robotics because swarm robotics does not require pre-existing man-made infrastructure. It can operate without telecommunications infrastructure like GPS or Wi-Fi.”
The real-life applications for such technology still evoke scenes from Hollywood.
Ferrante imagines that someday, scientists will be able to send a swarm of robots into space to explore another planet, maybe building a base like termites do on Earth. Robotic swarms could be used, for example, to scan and map a vast indoor environment, like a decommissioned nuclear power plant, or to find the source of a radiation leak.
Ferrante said it’s possible defense departments around the world are already using some form of this technology, but there is no way to know for sure, because this is no research that has been published.
So, how close is humanity to fast-moving, smart swarms of robots that can complete any task? Ferrante said the ideal system does not yet exist. A perfectly working swarm of robots would be able to do more than one task and complete an entire mission, whatever it is. However, according to Ferrante, current research focuses on pieces of the task, while integrating them into complete real-world solutions remains challenging.
“Eventually, you get to something that solves a problem, but that takes a lot of time,” Ferrante said.
Ferrante is excited to be at NYUAD, as it is the perfect place to conduct his futuristic, visionary work.
“The culture is amazing,” Ferrante said, “the culture of the colleagues, the culture of the students. The type of support this institution has and the clear plan to become self-sustained. It’s going in the right direction.”
