Eduardo Montijano - Distributed Consensus with Visual Perception in Multi-Robot Systems

Distributed Consensus with Visual Perception in Multi-Robot Systems — read online for free the complete book (whole text) full work

Below is the text of the book, divided by pages. System saving the place of the last page read, allows you to conveniently read the book "Distributed Consensus with Visual Perception in Multi-Robot Systems" online for free, without having to search again every time where you left off. Put a bookmark, and you can go to the page where you finished reading at any time.

Light

Font size:

↓

↑

Georgia Tahoma Arial Verdana

Reset

Interval:

↓

↑

Bookmark:

Make

1234567...82

Springer International Publishing Switzerland 2015

Eduardo Montijano and Carlos Sags Distributed Consensus with Visual Perception in Multi-Robot Systems 10.1007/978-3-319-15699-6_1

1. Introduction

Eduardo Montijano 1

(1)

Centro Universitario de la Defensa, Academia General Militar, Zaragoza, Spain

(2)

Departamento de Informtica e Ingeniera de Sistemas, Universidad de Zaragoza, Zaragoza, Spain

Eduardo Montijano (Corresponding author)

Email:

Carlos Sags

Email:

The first chapter of this book is devoted to discuss the concept of a multi-robot system and its applications in our society. Then, we analyze the implications of using vision sensors in such systems and the problems that need to be addressed in such a setup. Finally, the chapter presents a detailed outline of the rest of the chapters of the book, including a high level description of the particular topics treated inside them.

1.1 Motivation

The key ideas of the book go around multi-robot systems and multi-robot perception. The use of vision systems onboard the robots and the consensus when using these sensors is a challenging problem of recognized interest.

1.1.1 Multi-robot Systems

The idea of autonomous robots helping humans in the execution of different tasks is nowadays a reality. During the last decade, there has been an enormous advance in technology that has made it possible for individual robots to perform hard tasks with high precision [123]. From simple repetitive tasks like cleaning a house to more sophisticated operations such as mine clearance, or cleanup and recovery of oil spills, robots have become an important element in our lives, Fig.. However, as happens with humans, cooperative work seems more attractive and produces better results than individual work.

Fig. 1.1

Robots have become an important element in our lives. A vacuum robot cleaning a house ( left ) and a robot handling an explosive ( right ). Right figure reproduced with permission of the authors [103]

Fig. 1.2

Brigade of firefighters acting cooperatively. Figure reproduced with permission of the authors [103]

Fig. 1.3

Team of robots

Let us consider for example a brigade of firefighters entering a house on fire, Fig.. It is obvious that a team of firefighters will perform all the tasks faster and better than a single firefighter. In order to success, the team requires a common knowledge about the house, for example, the state of the fire in the different rooms, where possible survivors are located, or where the exits of the building are.

If instead of a firefighters brigade the team is integrated by robots (Fig. ) introduces new challenges that designers of these systems must face.

One way to deal with multiple robots is by using a central computer that receives and processes all the information about all the robots. Centralized approaches use a specific computer, usually called server, which optimally takes the decisions and coordinates the team. These approaches are easier to design and to implement, however, they are hardly scalable and become unsuitable when the number of robots is large enough. Moreover, if the server has a failure, the task will not be accomplished, which gives the system less reliability.

Distributed approaches [68], on the other hand, are harder to design but in general perform better. These solutions assume that each robot has its own computer onboard and sends the information about the scene to a subset of the robots, its neighbors. In this way, the team of robots is able to handle changes in the communication network, usually caused by changes in their positions or failures in their systems. As a consequence, the system is more reliable and independent of individual members, Fig..

Fig. 1.4

Centralized ( left ) versus distributed ( right ) solution. In a centralized setup, one robot gathers information about the whole network and makes the decision. In a distributed scenario all the robots play the same role and exchange messages with all their neighbors. As a consequence, the system is more reliable and independent of individual members

Researchers of multi-robot systems have turned to nature on many occasions to find ideas to solve different coordination problems in a distributed way. This occurs because nature contains perfect examples of distributed cooperation that could be replicated by a team of robots. Each animal can be seen as an autonomous element interacting with the environment. But at the same time, there is cooperation and interaction with other animals of the same species. In some cases each member of the team is assigned a different role, like ant colonies and swarms of bees, where explorers look for food, builders construct new passages in the colony, and the queen is in charge of breeding new offspring.

Other species do not distinguish between different members, every individual behaves equally with the others, following a set of simple rules that lead to emergent behaviors. This is the case, for example, of a fish shoal or a flock of birds moving in formation. There is no leader conducting everybody, each animal individually makes decisions about the direction to move by just looking at near mates. However, the whole group of animals ends up moving in the same direction, forming a specific pattern, Fig..

Fig. 1.5

A fish shoal ( left ) and a flock of birds ( right ) moving in formation. Each individual moves only considering the motion of those animals close to it. However, the full group of animals presents an emergent behavior, all moving in the same direction

This second scenario is very appealing in multi-robot systems [53], to begin with, because all the robots play the same role in the task. This simplifies the task of designing the system because the same algorithm is executed by all the robots without differences. Secondly, because the robots do not require to have full knowledge about the information perceived by all the robots, it is just enough to exchange information with the robots nearby. Finally, because a global behavior is still achieved even when everything is done locally, the system is able to reach some global behavior, e.g., all the robots move in the same direction.

1.1.2 Distributed Consensus and Vision Sensors

In this framework, one of the most important emergent behaviors that can be asked in a multi-robot system is the consensus between all the robots. The idea of reaching a consensus has great interest in many cooperative tasks. Generally speaking, we say that the team of robots has reached a consensus when all the robots agree about some data value which is of interest to the team. The essence of these data will depend on the specific application context. For example, the data can be the orientations or the positions of all the robots, and the consensus can be used to make the team of robots move in the same direction, attitude consensus [53, 92, 112], like in Fig. , or gather together at a specific point, rendezvous [27].

Next page

1234567...82

Light

Font size:

↓

↑

Georgia Tahoma Arial Verdana

Reset

Interval:

↓

↑

Bookmark:

Make

Similar books «Distributed Consensus with Visual Perception in Multi-Robot Systems»

Look at similar books to Distributed Consensus with Visual Perception in Multi-Robot Systems. We have selected literature similar in name and meaning in the hope of providing readers with more options to find new, interesting, not yet read works.

Andrew Tanenbaum

Distributed operating systems

Arup Kumar Sadhu

Multi-Agent Coordination (IEEE Press)

MagdiSadek Mahmoud

Control Design of Multiagent Discrete-Time Systems

Imran Bashir

Blockchain Consensus: An Introduction to Classical, Blockchain, and Quantum Consensus Protocols

Richard Astbury

Microsoft Orleans for Developers: Build Cloud-Native, High-Scale, Distributed Systems in .NET Using Orleans

Joseph Thachil George

Introducing Blockchain Applications: Understand and Develop Blockchain Applications Through Distributed Systems

Ziyang Meng

Modelling, Analysis, and Control of Networked Dynamical Systems

Dimos Raptis

Distributed Systems for practitioners

Thomas Hunter Ii

Distributed Systems with Node.js

Gibb Randy

Aviation visual perception: research, misperception and mishaps

Arslan Munir

Modeling and Optimization of Parallel and Distributed Embedded Systems

Ronald C. Arkin

Behavior-Based Robotics

Discussion, reviews of the book Distributed Consensus with Visual Perception in Multi-Robot Systems and just readers' own opinions. Leave your comments, write what you think about the work, its meaning or the main characters. Specify what exactly you liked and what you didn't like, and why you think so.