Robotics and Autonomous Systems 58 (2010) 1292–1305 Contents lists available at ScienceDirect Robotics and Autonomous Systems journal homepage: www.elsevier.com/locate/robot Linked multi-component mobile robots: Modeling, simulation and control Z. Echegoyen, I. Villaverde, R. Moreno, M. Graña ∗ , A. d’Anjou Computational Intelligence Group, Dept. CCIA, UPV/EHU, Apdo. 649, E-20080 San Sebastian, Spain article info Article history: Available online 25 September 2010 Keywords: Multi-robot systems Modeling Simulation Visual servoing Fuzzy control abstract The Linked Multi-Component Robotic Systems (L-MCRS) consists of a group of mobile robots carrying a passive uni-dimensional object (a hose or a wire). It is a recently identified unexplored and unexploited category of multi-robot systems. In this paper we report the first effort on the modeling, control and visual servoing of L-MCRS. Modeling has been tackled from geometrical and dynamical points of view. The passive element is modeled by splines, and the dynamical modeling is achieved by the appropriate extension of Geometrically Exact Dynamic Splines (GEDS). The system’s modeling allows realistic simulation, which can be used as a test bed for the evaluation of control strategies. In this paper we evaluate two such control strategies: a baseline global controller, and a fuzzy local controller based on the observation of the hose segment between two robots. Finally, we have performed physical experiments on a team of robots carrying a wire under a visual servoing scheme that provides the perceptual information about the hose for the fuzzy local controller. Visual servoing robust image segmentation is grounded in the Dichromatic Reflection Model (DRM). © 2010 Elsevier B.V. All rights reserved. 1. Introduction and motivation Multi-component robotic systems (MCRS), or multi-robot sys- tems, have been proposed in several application domains as a way to either perform a task that cannot be done by a single robot, or improve the efficiency (time or cost) of its realization even if it can be done by a single robot. Collaborative perception, morphological self-organization to overcome obstacles [1] or to establish commu- nication links, are among the tasks performed by MCRS. In general, outside the industrial deterministic realm, MCRS have been pro- posed as a way to overcome the difficulties imposed by unstruc- tured and non-stationary environments. There are several review works in the literature giving different categorizations of MCRS [2–6] focusing on different aspects of the multi-robot systems (communication topology, cooperation, geometrical restriction, learning mechanisms and others). A categorization of MCRS in terms of the degree of physical coupling among the individual robots is presented in [6]. That paper contains a discussion of sev- eral aspects of MCRS: their morphology, the tasks they have to per- form and the environment in which they have to be carried out, the control of the system and the perception used to obtain feedback from the actions taken and their effects. However, for our purposes the main relevant idea of [6] is the identification of three main types of MCRS according to the individual’s coupling degree: ∗ Corresponding author. E-mail address: manuel.grana@ehu.es (M. Graña). URL: http://www.ehu.es/ccwintco (M. Graña). • The Distributed MCRS (D-MCRS) which corresponds to groups of (mobile) robots without physical connection. Typical tasks for this kind of systems are collaborative perception and mapping of an environment, cooperative transportation of collections of items, etc. • The Modular MCRS (M-MCRS) which corresponds to groups of modular robotic elements attached with rigid (strong, fixed) links to assume a given morphology which can be task-depe- ndent. Typical tasks are self-organization for obstacle avoid- ance and environment adaptation, cooperative transportation of large rigid items. • The Linked MCRS (L-MCSRS) which corresponds to groups of autonomous (mobile) robots linked through a passive non- rigid linking element. This passive element introduces dynamic problems and restrictions that may greatly influence the con- trol of the whole robot ensemble. A typical task would be the manipulation and transportation of a hose or wire. Both D-MCRS and M-MCRS have been dealt with extensively in the literature, however we find that the L-MCRS is a new category, not previously identified in the literature. To illustrate this point, let us consider the task of self-perception, the ability to measure the configuration of the system. The M-MCRS may sense its configuration through the state of its rigid joints connecting modules. The D-MCRS may estimate it from the information gathered by the individual robots independently. The L-MCRS, however, needs to perform both the estimation of the individual robot states and the state of the passive linking element. While the individual robot states can be estimated directly through odometry, the passive linking element state can only be measured through remote (visual) observation. 0921-8890/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.robot.2010.08.008