Swarm robotics: Why one engineering professor looks to insects for robotic inspiration
Dr. Andrew Vardy, an associate professor who is joint appointed with the Faculty of Engineering and Applied Science’s Department of Electrical and Computer Engineering and the Faculty of Science’s Department of Computer Science, is interested in ants and bees.
While that may seem a little odd for an engineering professor, Dr. Vardy is quick to point out that engineers can learn a lot from these social insects.
“They are interesting as inspirations to engineers because of how their colonies function,” said Dr. Vardy.
Social insect colonies are fault-tolerant, which means they can survive even if some individual members get sick or die; they are scalable, which means they have the capacity to continue even while the number of members go up or down; and they are adaptable, which means members can change their behavior in response to a changing environment.
“Current robot technology is generally not fault-tolerant, scalable or adaptive,” said Dr. Vardy. “So it seems that robotics can learn a lot from the social insects.”
Dr. Vardy’s main research area is called swarm robotics.
Swarm robotics concerns the development of multi-robot systems inspired by the social insects, such as bees and ants, which live together in colonies and build incredibly sophisticated structures given their limited individual capabilities.
One recent outcome of Dr. Vardy’s research is an improved algorithm for swarms of simple robots to sort objects.
The robots are deployed in an environment containing a random distribution of coloured pucks.
“The robots have cameras to detect and classify the pucks, as well as to sense other robots, and apply relatively simple control laws to sort the pucks by colour,” said Dr. Vardy. “For example, a simple controller picks up isolated pucks and puts them down next to clusters of pucks of the same colour. But, we can go further by remembering locations and programming the robots to return to previously visited clusters of pucks.”
Dr. Vardy’s research uses spatial memory.
In contrast to previous work, these robots can remember places within their environment and return to them, which is similar to bees and ants.
The intended application is recycling.
Instead of a large dedicated recycling facility, such as the one in Robin Hood Bay, swarms of robots can be applied to sort out different types of plastic and metal containers.
“The inspiration from insects leads to a scalable system, which means this concept can be applied in large or small communities,” said Dr. Vardy. “This is quite relevant in Newfoundland and Labrador where there are many small, remote communities. The remoteness of these communities is often a barrier to the use of traditional recycling methods.”
Dr. Vardy has students working in various facets of swarm robotics.
“One of my students is programming robots to form into moving patterns, sweeping across an environment, such as the seabed, and capturing images,” said Dr. Vardy. “Another student is looking at how a single human can control a whole swarm of semi-autonomous robots.”
The main goal of Dr. Vardy’s research is to program swarms of robots to modify the environment.
“We have robots sorting pucks, and now we would like to have robots building structures,” said Dr. Vardy. “One of my students is programming robots to re-shape terrain by driving across it. We want to take inspiration from insects that modify their environments to help support their colonies. Ultimately, we want to develop robot swarms that can help us clean up and re-shape our environment.”
More details on Dr. Vardy’s research are available at bots.cs.mun.ca.