In nature, we observe that many insects like bees and ants work in unison . Their efficiency is observed to be higher than any individual. In a similar way, a swarm of robots is more efficient than a single robot, due to less hardware and small size. Swarm robotics is an emerging branch that is interdisciplinary. It covers areas such as mechatronics, communication, and artificial intelligence. The projects aim to implement collective action of robots to accomplish small tasks. A decentralized approach has been used. A suitable algorithm has been designed that implements these collective actions. The heart of a swarm agent is an Atmega328p micro-controller. To implement the inter-agent communication, IR medium of communication is used with NEC protocol. In a manual mode of operation, the user can relay commands
to a single robot using Zig-Bee communication. The three basic tasks are implemented on the prototype of swarm robots. They are namely; follow the leader in which agents follow the behavior of a designated agent, cluster formation in which all agents form a predefined shape or pattern. These tasks are supplemented by localization in their environment i.e. mapping of their immediate surrounding environment.
Nature has always been the source of inspiration for mankind. No matter how much progress we make, it's always insignificant when compared with natural intelligence. By keeping this in mind, the field of robotics has started its journey on a newly emerging field where there is the union of biology and many other branches. Swarm robotics is such a branch. Ant colonies, bee swarms, and schools of fish are the few examples that demonstrate the power of swarms. In the same way, a group of robots is much more effective than a single agent.
The tasks performed by the robot may not always be the same. Hence, there is always variance in the efficiency of the system. However, in a swarm, only the necessary agents are called to accomplish a task thus resulting in maximum efficiency of the system.
The various advantages of swarm robotics that make this field an interesting one are scalability, robustness, cost-effectiveness, energy efficiency and the most important one is intelligence.
As seen in nature, the colony size of ants may vary but their functions remain unchanged. In the same way, once a swarm of robots has been developed, the number of agents can be varied in any number. The intelligently designed algorithms are completely independent of the number of agents. The swarm gets only more effective when the number of agents is sufficiently high.
The efficient working of any system is completely dependent on the successful execution of tasks by every sub-system embedded in it. So a malfunction of any of the parts results in the collapse of the entire system. However, in swarms, the working of the system is independent of every agent. In case an agent fails, the communication is bypassed and the system remains intact.
The agents in swam are optimized in terms of size, power, and features. This results in an extremely simple design in terms of hardware electronics. The sheer intelligence is the result of algorithms and protocols. The inter-bot communication is carried out by Infrared LED which is also the main reason why Swarm Robotics becomes cost-effective.
Conclusion: Check out the project to understand more.