Introduction:

Ever wondered about swarm robotics? Glad you asked. Here is my overview of this interesting topic.

Interacting, knowing, and then reacting to the situation are some of the human ‘s greatest characteristics and these are the aspects that make us what we are. We are born to live in social society and we always know about us that since the creation of this planet we are the most well manned social creature known.

swarm robotics inspiration
Swarm Robotics Inspiration

Social culture and contact with each other to help accomplish a common purpose is seen not only in humans but also in other animals on this planet, such as a flock of birds or fish or bees, what they have in common is that they have a collective activity. This serves as an important motivation for swarm robotics development. We also see that as the birds migrate they are often in a group headed by the leading member of their group and they all follow them and their group is built in a specific geometrical form given the fact that the birds have no understanding of the shapes and figures and the group is rendered so that the senior members of the group are on the boundaries while the young or the newborn is on the boundaries.

The same characteristics are found in fire ants, these ants are a bit different from other species of the ants and are especially known for their group behavior, they build together, they eat together and they defend their colonies from the preys together, basically they know they can achieve more when they are in a group. A recent study was being conducted on the group behavior of these ants in which it was found that they were capable of making strong structures whenever needed, such as when needed to create a small bridge to crossover.

For all their limitations, the mutual action of these social animals and insects makes them accomplish more. Researchers have shown that no participation or sophisticated information is required by individuals of these groups to generate these complex behaviors. In social insects, the individuals of animals and birds are not informed of the colony ‘s global status. Swarm information is spread among all agents, where a person can not accomplish his or her task without the rest of the swarm. How if it is possible to put this mutual sensing into Robots?

Swarm robotics is a new-age technology, swarm-bots communicate in a decentralized way locally with each other and the environment, to achieve the ultimate goal by self-organization.

To understand Swarm robotics, we need to understand swarn robots the following criteria differentiates swarm robots from simple robots:

  • Swarm robots are autonomous robots and they can operate in real environments.
  • It is necessary to be homogeneous in the program and algorithm. They can be different types of robots in a swarm but algorithms and communication protocols must be the same.
  • They are many in number and must be cooperative.
  • They work together to solve problems.
  • The robots do have capacities for contextual connectivity and sensing. It guarantees the distribution of information, such that scalability is one of device properties.

                                                                                      swarm robots

                                                                                      Figure 2: Swarm Robots

Motivation and Inspiration:

Multi-robotic systems maintain some of the social insect ‘s characteristics such as robustness, the robot swarm can work even if some of the individuals fail, or there are disruptions in the surrounding environment; flexibility, the swarm can create different solutions for different tasks and can change each robot ‘s role depending on the need for moment. Scalability, robot swarm is capable of operating in different group sizes, from a few individuals to thousands.ant swarm

 

Characteristic:

characteristics of group mechanics

As said, Swarm robotics involves simple swarm robots acquire a social insect characteristic which is described as follows:

  1. This robot must be autonomous and capable of sensing and behaving in real surroundings.
  2. The number of robots in a swarm must be sufficiently high to support every single task as a group to be performed.
  3. Homogeneity in the swarm should be present, there will be different classes in the swarm but they should not be too many.
  4. A single swarm robot must be incapable and inefficient with respect to its main objective, that is to say, they must work together to succeed and improve the results.
  5. All the robots are expected to only have local sensing and communication capabilities with the swarm ‘s neighboring partner, this ensures swarm coordination is distributed and scalability is one of the system ‘s assets.

Swarm Intelligence:

swarm intelligence

Swarm Intelligence (SI) can be defined as a relatively new branch of Artificial Intelligence that serves to model social swarms’ collective behavior in nature, such as ant colonies. It can be defined as “The emerging collective intelligence of groups of simple agents”. Swarm intelligence (SI) is the collective behavior of natural, or artificial, localized, self-learning systems. The principle is used in artificial intelligence research. Gerardo Beni and Jing Wang introduced the expression in 1989, in the context of cellular robotic systems and as a set of algorithms for controlling robotic swarms in the global optimization framework. Swarm Intelligence concepts have been widely applied in a number of problem areas including task optimization problems, identifying optimal paths, routing, structural optimization, and image and data processing. Sometimes, the impetus comes from the essence of the social insects, how they work and interact in various contexts. The agents obey very simple rules which lead to the development of “intelligent” global behavior, unknown to the individual agents in which there is no centralized control mechanism directing how individual agents will act. Furthermore, the research in swarm intelligence has led to the development of the different swarm modes which operate on the principle of social insect collaboration. The swarm intelligence models that can also be considered natural swarm-inspired computer models. Examples of swarm intelligence models are: Ant Colony Optimization, Particle Swarm Optimization, Artificial Bee Colony, Cat Swarm Optimization and Artificial Immune Network.

 

cat group optimization

                                                                              Figure 5:Cat Swarm Optimization

Swarm robotics system-level:

A swarm robotic system-level operation will demonstrate three practical properties which are found in natural swarms and exist as attractive properties in multi-robot systems:

Flexibility:

A swarm entity would be able to organize their actions to tackle activities of specific character. For example, individuals in an ant colony can jointly find the shortest path to a food source or bring a large prey with the use of various organizing strategies.

Robustness:

The swarm robotic system should be able to run through environmental disruptions or its individuals malfunctioning. A variety of reasons behind the robustness of their activity can be found in social insects. Firstly, swarms are essentially redundant systems; an agent will automatically compensate for the lack of another agent. Second, control is dispersed and it is impossible that the failure of a single section of the swarm can disrupt its activity. Second, the individuals making up the swarm are fairly easy, making them less vulnerable to failure. Fourthly, sensing is distributed; thus, the system is stable to local environmental perturbations.

Scalability:

The swarm will be capable of working across a wide range of group sizes and serving vast numbers of people without dramatically affecting results. This is, the management mechanisms and techniques for swarm robotic systems to be developed will ensure swarm activity under differing swarm sizes.

Experimental Platforms:

Different experimental platforms used in the most applicable literature swarm-robotic studies, including robotic platforms and simulators, are listed here.

Platforms:

Name Size (mm) (diam.) Actuators Sensors Communications Relative positioning system Development
Khepera 55 Wheeled
(differential drive)
8 IR RS232
Wired link
Research
Commercial
Khepera III 120 Wheeled
(differential drive)
11 IR
5 ultra sound
WIFI and Bluetooth Expansion
(IR based)
Research Commercial
e-puck 75 Wheeled
(differential drive)
11 infrared (IR)
Contact ring
Color camera
Bluetooth Expansion
(IR based)
Open-source
Commercial
Research and Edu.
Alice 2 0 × 2 0 Wheeled
(differential drive)
IR proximity and light
Linear camera
Radio
(115 kbit/s)
Research
Noncommercial
Jasmine 2 3 × 2 3 Wheeled
(differential drive)
8 IR IR Integrated
(IR based)
Open-source
Research
I-Swarm
robot
3 × 3 Micro legged a a Research
Non commercial
S-Bot 120 Wheeled
(differential drive)
2 Grippers
15 Proximity
Omni Camera
Microphone, Temp.
WIFI Camera based Research
Non commercial
Kobot 120 Wheeled
(differential drive)
8 IR
Colour camera
Zigbee Integrated
(IR based)
Research
Noncommercial
SwarmBot 1 2 7 × 1 2 7 Wheeled
(differential drive)
IR, light sensors
Contact, camera
IR based
(Local)
Integrated
(IR based)
Research
Noncommercial

Simulators:

There are several available handheld robotic simulators that can be used in multi-robotic experiments, and more specifically in swarm-robotic experiments. They vary in their technological aspects as well as in the license and rate. We outline them in the following section, along with feedback on their use in swarm-robotic applications.

Webots: Webots is a practical, commercial smartphone simulator allowing multi-robot simulation, with versions of actual robots already developed. This is 3D, simulating collisions and mechanics. According to our experience its efficiency declines very quickly when operating with more than 100 robots, making it impossible to replicate with a large number of robots.

Microsoft Robotics Studio: MRS is a Microsoft Corporation-developed simulator. This needs such a computer that runs Windows.Know more about Microsoft Robotics studio here

SwarmBot3D: It is Designed by the i-Robot company for research.Know more about Swarmbot3d

Basic Behavior and Tasks in Swarm Robots:

Within this section, a selection of the most influential theoretical works is represented within Swarm Robotics. The numerous experimental findings are categorized, grouping them according to the activities or actions conducted by the swarms. Many of the actions, such as grouping and joint action, are very simple and for more complex activities represent a prior stage. We are portrayed in highly complicated order.

Dispersion:

The aim of dispersion is to disperse the robots in space to cover as much area as possible, usually without losing the interconnectivity. When scattered, the swarm will act as a global sensor but also as an exploratory tool. Dispersion has been researched both using actual robots and in modeling by numerous researchers. propose a possible field algorithm for the robotics implementation, in which barriers and other robotics repel robots. The method is generalized and needs no centralized localization, which leads to a scalable solution. The research is performed in simulation only.

Aggregation:

Aggregation is one of the fundamental behaviors of swarms in nature and is observed in organisms ranging from bacteria to social insects and mammals. Aggregation helps species deter predators, withstand aggressive environments in the environment, and attract mates. Any of the activities of aggregation are considered to be encouraged by environmental clues; flies use light and temperature, and sowbugs use moisture to accumulate. Many aggregations are therefore self-organized. The gathering of cockroaches, young penguins, and fish schools does not use these prompts but instead is the product of a joint decision occurring.

Collective Movement:

Collective action is the question of how to organize and cohesively move a community of robots together as a collective. This can also act as a foundational action for more complex activities. It can be classified into two types: flocking and formations. In the former robots have to maintain between them predetermined positions and orientations. On the other hand, the numerical locations of robots are not specifically applied at flocking.

Task Allocation:

Some robots have already reported assignments and all of them will attend to them at the same time. The first algorithm is based on a gossip communication scheme, and it has better performance than the other one, but it may be less scalable due to the limited robustness to packet loss. The second is simple and receptive, based on light-signal interaction.

Still not fascinated by the power of  swarm robots ??watch the video below ..

Want to publish your first robotic blog ?

Read the community guidelines and starting writing today !!!
TR Community Guidelines



No Thoughts on Swarm Robotics: An Easy Overview

Leave A Comment