The Intersection Between Swarm Intelligence and Robots
Do you know what swarm intelligence is? Fear not, this article will sharpen your mind on the world of collective behaviors, and how I transformed it into robots.
What is Swarm Intelligence?
Daily, tiny insects such as ants are seen around you every day. Them alone seem almost helpless; however, in groups, you see incredible tasks being done. While I started to notice these creatures more and more often, I became more curious about how these creatures were able to communicate with one another. Just noticing these animals sparked a thought in my brain: what is this group of intelligence, and how does it work?
Throughout my research, I learned various things about these groups of intelligence. Swarm intelligence is It is the collective behavior of self-organized systems, inspired by ants building anthills, termites building mud structures, and birds migrating. All swarm intelligence groups, such as ants and bees, aren’t knowledgeable alone. Ants, for instance, may do only simple things like sniff fermented trails if food is necessary. But what’s really mind-blowing about these microscopic creatures is that they, together, can form a superorganism that can do breath-taking tasks that are impossible for one to do alone.
I hypothesized a couple of rules for my robots to mimic the behaviors of these collective behaviors:
- Robots must be aware of their abilities and surroundings.
- Robots must work in solidarity, meaning that they must work without a central command and leader.
- Robots’ work is never affected, even when members are added and removed dynamically.
I designed and programmed multiple robots to mimic this incredible behavior; just like bees and ants, all robots were built to be Small, Simple, Similar, and Self-Contained. These robots are programmed to work based on local information about their neighbors. Similar to the animal swarms organizing behaviors, the robots are trained to accomplish complex tasks working alongside the robots.
Components I Used
The swarm network is designed with three main components: the Particle Argon, the Particle Xenon, and Thread®. What are these things, you may ask? Thread®, to begin with, is an open-source mesh networking protocol. Mesh networking is based on the universally-supported Internet Protocol (IP) and is built using open and proven standards. In other words, Thread® is the foundation that the mesh network I created is built upon. Originally created by Nest, my mesh network can use it as an open-source implementation of the communication between robots.
The Particle Argon works as the gateway for the mesh network, which uses the Particle Xenons to run each robot in the swarm. Each robot uses its knowledge to detect the presence and neighbors, giving it a self-healing and locally autonomous nature.
Another essential component I used is the Crickit Chip. Like the sport, or the bug? No, silly. The Crickit Chip is used for the motor ports and Capacitive Touch of my device. The motor ports allow the movement of my robots, whereas Capacitive Touch allows my robots to detect the presence of one another. Capacitive Touch, in other words, basically works like your phone. There are multiple layers, with the inner and outer layers being able to conduct electricity. Since your finger works almost like a conductor, depending on the area you touch on your phone, it is able to complete different tasks.
The robot’s foundation is built using a 3D printed robot chassis from my local library. Each robot was built with the same components, chassis, and Particle Xenon chip, showing the ultimate resemblance of each robot, similar to those behaviors of ants and bees.
Demonstrations of My Swarm Principles
Now, each of the three swarm principles had a demonstration to showcase how my robots are able to successfully demonstrate that principle.
Here are the three demonstrations:
Swarm Principle One: In this demonstration, the robots demonstrate collective behaviors similar to ants and bees, by working together and communicating in a synchronized manner. The robots, standing in a straight line, are able to move forward, backward, and turn, in unison with their neighbors. A message is sent through the Particle app to any robot in the group, and it can transfer that message to the entire swarm, in a methodical and intelligent approach. This demonstrates the principle that the robots must work in complete synchronization; even when the members of the swarm are added or removed, the behaviors of the entire group of robots aren’t affected, nonetheless of how the amount of robots is adjusted. This principle can allow the synchronization and control of the swarm, which allows us to command specific tasks easily.
Swarm Principle Two: In this demonstration, the robots work together to search for a light source, which in ants resemblance, would be a “food source” of some sort. The robots are each equipped with copper tentacles, which use the Capacitive Touch sensors to detect the presence of one another. The boundary that the robots are placed in stimulates a repelling force to the tentacles as well. The robots use light sensors, attached at the bottom of each chassis, to be able to detect the light source. When the light source is detected by a single robot, it sends a message through the Particle mesh network, notifying the rest of the swarm that the source had been found. The robots then stop, the LED lights on the robots are able to differentiate which robot found the light source, and which didn’t. This demonstration supports the principle that all robots must be aware of their surroundings and their abilities; this can be helpful in the need of search and rescue operations, in which multiple robots can communicate to detect a specific object and/or human.
Swarm Principle Three: In this demonstration, the robots are asked to push a heavy object. In the beginning, one robot alone attempts to push the object. When the robot isn’t successful, it sends an Alert message to the other robots in the swarm network, and, in a few seconds, all the available robots help the first push the object together. This demonstration supports the principle stating that all robots must work in solidarity, meaning that the work put amongst themselves must be done without the central command of a leader. We can use this integrated intelligence within the swarm in construction, health care, warehousing, and many other industrial sectors.
My Results and Conclusions
Just as this project demonstrates, swarm intelligence has a wide range of applications ranging from industrial, medical to military sectors, construction to rescue operations working together in areas where it would be too dangerous or impractical for humans and also where there is no infrastructure in place for communication.
Swarm robots are extremely important in agricultural applications where the swarm of aerial vehicles can record and report the health of each crop and inform the farmer to take the necessary steps to avoid any loss. The swarm nodes can interact with each other without connecting to any central command. Robot swarms have a great necessity in improving manufacturing processes and workplace safety. They can be deployed to help to monitor water resources, detect hazardous events such as chemical leakages. Swarm intelligence can be used extensively for high-performance parallel processing, autonomous routing, and data processing. Swarms can be used for the treatment of many cancers, where we can send in very small robots with their collective intelligence that can find the cancer site and attack very specifically.
My future goal is to increase the number of robots in the swarm to investigate further collective behavior and intelligence. Rather than adding more sensors to each robot, I would like to modify software instructions to connect to an Artificial Intelligence source and use the swam intelligence as an add on to perform more complicated tasks. I also would like to work on further reducing the size of each robot in the swarm, this can not only help reduce the cost but also will help increase the density of the swam to perform tasks faster and better.
In conclusion, the project’s main purpose is to find ways to make people’s lives better with the above-mentioned goals using swarm robots and their collective intelligence just as this project just exhibited the robots group intelligence. Again, Swarm Intelligence is where each member autonomously offers its abilities and their collective behaviors for the greater good. I hope my project can influence the world to use swarms in healthcare, construction, warehousing, and other fields!
