Swarmanoid: a novel concept for the

study of heterogeneous robotic swarms

M. Dorigo, D. Floreano, L. M. Gambardella, F. Mondada, S. Nolfi,
T. Baaboura, M. Birattari, M. Bonani, M. Brambilla, A. Brutschy, D. Burnier, A. Campo,
A. L. Christensen, A. Decugnière, G. Di Caro, F. Ducatelle, E. Ferrante, A. Förster,

J. Martinez Gonzales, J. Guzzi, V. Longchamp, S. Magnenat, N. Mathews,

M. Montes de Oca, R. O’Grady, C. Pinciroli, G. Pini, P. Rétornaz, J. Roberts, V. Sperati,

T. Stirling, A. Stranieri, T. Stützle, V. Trianni, E. Tuci, A. E. Turgut, and F. Vaussard.

IRIDIA – Technical Report Series

Technical Report No.
TR/IRIDIA/2011-014
July 2011
IRIDIA – Technical Report Series
ISSN 1781-3794

Published by:
IRIDIA, Institut de Recherches Interdisciplinaires
et de Développements en Intelligence Artificielle
Université Libre de Bruxelles
Av F. D. Roosevelt 50, CP 194/6
1050 Bruxelles, Belgium

Technical report number TR/IRIDIA/2011-014

The information provided is the sole responsibility of the authors and does not necessarily
reflect the opinion of the members of IRIDIA. The authors take full responsibility for
any copyright breaches that may result from publication of this paper in the IRIDIA –
Technical Report Series. IRIDIA is not responsible for any use that might be made of
data appearing in this publication.
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 1

Swarmanoid: a novel concept for the study of

heterogeneous robotic swarms
Marco Dorigo, Dario Floreano, Luca Maria Gambardella, Francesco Mondada, Stefano Nolfi,
Tarek Baaboura, Mauro Birattari, Michael Bonani, Manuele Brambilla, Arne Brutschy, Daniel Burnier,
Alexandre Campo, Anders Lyhne Christensen, Antal Decugnière, Gianni Di Caro, Frederick Ducatelle,
Eliseo Ferrante, Alexander Förster, Javier Martinez Gonzales, Jerome Guzzi, Valentin Longchamp,
Stéphane Magnenat, Nithin Mathews, Marco Montes de Oca, Rehan O’Grady, Carlo Pinciroli, Giovanni Pini,
Philippe Rétornaz, James Roberts, Valerio Sperati, Timothy Stirling, Alessandro Stranieri, Thomas Stützle,
Vito Trianni, Elio Tuci, Ali Emre Turgut, and Florian Vaussard.

Swarm robotics systems are characterised by decentralised The swarmanoid exploits the heterogeneity and comple-
control, limited communication between robots, use of local mentarity of its constituent robot types to carry out complex
information and emergence of global behaviour. Such systems tasks in large, 3-dimensional, man-made environments.1 The
have shown their potential for flexibility and robustness [1], system has no centralised control and relies on continued
[2], [3]. However, existing swarm robotics systems are by local and non-local interactions to produce collective self-
in large still limited to displaying simple proof-of-concept organised behaviour. The swarmanoid architecture provides
behaviours under laboratory conditions. It is our contention properties difficult or impossible to achieve with a more
that one of the factors holding back swarm robotics research conventional robotic system. Swarmanoid shares the strengths
is the almost universal insistence on homogeneous system of existing swarm systems. Robots of a particular type are
components. We believe that swarm robotics designers must directly interchangeable, providing robustness to failures and
embrace heterogeneity if they ever want swarm robotics systems external disturbances. However, swarmanoid’s heterogeneous
to approach the complexity required of real world systems. nature gives it a flexibility that previous swarm systems
To date, swarm robotics systems have almost exclusively cannot match. Different sensing and actuating modalities of
comprised physically and behaviourally undifferentiated agents. its heterogeneous components can be combined to cope with
This design decision has probably resulted from the largely a wide range of conditions and tasks. The swarmanoid even
homogeneous nature of the existing models that describe self- features dynamic self-reconfigurability: groups of robots can
organising natural systems. These models serve as inspiration get together on a by-need basis to locally form ad-hoc coalitions
for swarm robotics system designers, but are often highly or integrated structures (by connecting to each other) that can
abstract simplifications of natural systems. Selected dynamics perform more complex tasks. Thanks to the heterogeneity of
of the systems under study are shown to emerge from the the robots in the swarm, these coalitions can flexibly integrate
interactions of identical system components, ignoring the a variety of skills.
heterogeneities (physical, spatial, functional, informational) To the best of our knowledge, the swarmanoid represents
that one can find in almost any natural system. the first attempt to study the integrated design, development
The field of swarm robotics currently lacks methods and and control of a heterogeneous swarm robotics system. In the
tools with which to study and leverage the heterogeneity that following sections, we first discuss the issues and challenges
is present in natural systems. To remedy this deficiency, we intrinsic to heterogeneous swarm robotics systems. We then
propose swarmanoid, an innovative swarm robotics system give an overview of the swarmanoid system. Finally, we
composed of three different robot types with complementary describe the experimental scenario devised to demonstrate the
skills: foot-bots are small autonomous robots specialised in capabilities of the swarmanoid.
moving on both even and uneven terrains, capable of self-
assembling and of transporting either objects or other robots; I. H ETEROGENEOUS ROBOTIC S WARMS : I SSUES AND
hand-bots are autonomous robots capable of climbing some C HALLENGES
vertical surfaces and manipulating small objects; eye-bots are Heterogeneous robotic swarms are characterised by the
autonomous flying robots which can attach to an indoor ceiling, morphological and/or behavioural diversity of their constituent
capable of analysing the environment from a privileged position robots. In a heterogeneous swarm robotics system, the need
to collectively gather information inaccessible to foot-bots and for physical and behavioural integration among the different
hand-bots (see Figure 1). hardware platforms results in a considerable amount of extra
complexity for the design and implementation of each different
IRIDIA, CoDE, Université Libre de Bruxelles, Belgium
LSRO, École Polytechnique Federale de Lausanne, Switzerland 1 Humanoid robots are usually assumed to be the most efficient robot type
LIS, École Polytechnique Federale de Lausanne, Switzerland for man-made environments. One of the goals of the Swarmanoid project was
IDSIA, USI-SUPSI, Manno-Lugano, Switzerland to refute this assumption. The term swarmanoid is, in fact, a compound of
ISTC, Consiglio Nazionale delle Ricerche, Roma, Italy swarm and humanoid.
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 2

(a) (b)

Fig. 1. The swarmanoid robots. (a) Three foot-bots are assembled around a hand-bot and are ready for collective transport. The hand-bot has no autonomous
mobility on the ground, and must be carried by foot-bots to the location where it can climb and grasp interesting objects. (b) An eye-bot attached to the ceiling
has a bird’s-eye view of the environment and can thus retrieve relevant information about the environment and communicate it to robots on the ground.

type of constituent robotic agent. This integration complexity assembly in homogeneous swarm robotics systems has proven
must be dealt with both in the hardware design, and at the challenging [5]. Designing and implementing self-assembly
level of behavioural control. capable hardware in a heterogeneous system is significantly
Robots within a heterogeneous swarm must be able to coop- more complex, as it involves managing potentially conflicting
erate. At the hardware level, this imposes the minimum require- requirements. The different robot types in a heterogeneous
ment that the various robot types have common communication swarm each have their own functionality requirements which
devices, and the sensory capabilities to mutually recognise impose constraints on morphology and on sensing and actuation
each other’s presence. Even this basic design requirement is capabilities. Self-assembly between heterogeneous robots, on
not trivial to realise. Robot communication devices are often the other hand, requires the different robot types to have a
tailored to a particular robot morphology and functionality. degree of compatibility in both their morphologies and their
Flying robots, for example, need communication devices that sensing and actuation capabilities.
are light and power-efficient, while for ground based robots Behavioural control is difficult to design for any swarm
higher performance devices that are heavier and consume more robotics system. Individual control rules must be found that
power may be appropriate. The challenge is thus to ensure that result in the desired collective behaviour. The complexity
functionally similar devices with very different design criteria resides in the indirect relationship between the robot’s proximal
can seamlessly interface with one another. level (i.e., the level of the individual controller, which deals
Swarm robotics systems also favour less direct interaction with the sensors, actuators and communication devices) and the
modalities. Stigmergic interactions, for example, are mediated swarm’s distal level (i.e., the overall organisation, which refers
by the environment [4], and have been shown to be effective in to the dynamics and self-organising properties of a complex
swarm systems. In a heterogeneous swarm, the difficulty is to heterogeneous robotic system).
ensure that actuation and sensing mechanisms on morphologi- In heterogeneous robotic swarms, the challenge is much
cally and functionally different robots manipulate and sense harder, as behavioural control must be able to integrate
the environment in a way that is sufficiently coherent to enable the different abilities of different robot types to work in
stigmergy. In fact, any form of non-symbolic communication synergy towards the achievement of a common goal. This
(e.g., visual communication using LEDs and a camera) requires integration must take into account both the specialisation and
a design effort to ensure a sufficient level of sensing and the complementarity of different robot types. Specialisation
actuation integration between robot types. means that each robot type has a specific set of tasks to which
Physical cooperation is often considered necessary in a it is particularly suited. Complementarity means that varied
swarm system to allow the swarm to overcome the physical actuation and sensing capabilities of the different robot types
limitations of single agents. An interesting possibility for allow them to work together in such a way that the whole is
physical interaction—often observed in biological systems—is more than the sum of its parts. In other words, the efficiency
self-assembly, that is, the ability of different individuals to of the heterogeneous swarm is greater than if the different
connect to one another forming a large physical structure. robot types worked independently in parallel without mutual
In robotics, this form of interaction can open the way to cooperation.
complex forms of cooperation. The implementation of self- To solve the behavioural control problem, it is necessary to
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 3

pursue a holistic approach, in which the possible interactions

between different robots in the heterogeneous swarm are taken
into account from the very beginning, before the development
and testing of the individual controllers. The choice of the
communication modality is crucial. Communication is an
essential aspect of any distributed robotic system, and can
take many different forms ranging from indirect stigmer-
gic interactions to networked structured communication. In
a heterogeneous swarm, it is necessary to consider both
intra- and inter-group coordination. To enable intra-group
coordination, it is necessary to develop interaction modalities
within homogeneous groups. To enable inter-group coordination
between groups composed of different robot types, the robots
must have a communication system that can convey useful
information to robots that experience the environment in a
different way. This requires the solution of novel problems,
such as defining shared attention mechanisms within and Fig. 2. The half credit card size i.MX31 main processor board.
between groups, or exploiting intra-group coordination and
communication as behavioural templates for the development
of inter-group coordination strategies: if a sub-group presents
some behavioural and/or communication traits, these could We designed and developed a common main processor board
represent a form of implicit communication for a different sub- for all the robot types. The board is based on a 533 MHz
group. Such implicit communication can be used to coordinate i.MX31 ARM 11 processor and features 128 MB of RAM,
different sub-parts of the heterogeneous robotic swarm. 64 MB of Flash, as well as a USB 2.0 host controller and
an energy and I/O companion chip (see Figure 2). The micro-
To support the development of robot behaviours for swarms
controllers are based on the DsPIC 33, as it provides good
of robots, simulation is a fundamental tool. Real-world ex-
computational power, includes fixed-point and DSP instructions
perimentation in swarm robotics is often impractical because
and has low power consumption.
of the necessity of testing behaviours with large numbers
of robots. Simulation of heterogeneous swarms poses further In order to access the different devices of the robot, we have
challenges, as the different robot types may have different developed a low-level software architecture called ASEBA [6]
simulation requirements. A simulation tool for heterogeneous that abstracts the peculiar features of the different robot modules
robots must, therefore, simultaneously offer scalability for and offers an easy-to-use tool for robotic experimentation.
increasing number of robots, and flexibility to support highly A SEBA is an event-based architecture consisting of a network
diverse robots designs. of processing units which communicate using asynchronous
messages called events. Usual read/write transactions from
the main processor to the micro-controllers are replaced
II. S WARMANOID T ECHNOLOGIES by events sent from any node to any other node on the
Research on the swarmanoid has been guided by the issues common communication bus. All nodes send events and react
broached in the previous section. As discussed, the various to incoming events. In our robots, the typical network is
constituent robot types must be able to interact, either physically formed by the main processor board and the various micro-
or through communication. We tackled the interaction problem controllers, which communicate through a Controller Area
from the outset by designing a set of common technologies to Network (CAN) bus. The micro-controllers correspond to the
provide a uniform hardware architecture. In this section, we different sensors and motor devices that are implemented on the
first describe these common technologies, and then detail the robot. Robot behaviours are based on the data that these sensor
hardware design of the three robotic platforms. Finally, we and motor devices provide. This data can be either processed
present the dedicated simulator that we developed. locally by the micro-controller, or can be communicated
through an asynchronous event. Asynchronous events are
implemented as messages that have an identifier and payload
A. Common Technologies data. By exchanging events and processing data both locally
All robots have a multi-processor architecture, consisting of and remotely, complex control structures can be implemented.
a main processor that takes care of CPU-intensive tasks such as The network of processing units can be extended through
vision and higher-level control, and several micro-controllers TCP-IP to any remote host. For development and debugging,
that take care of real-time sensor reading and actuator control. for example, an integrated development environment (IDE)
This design choice represents a clear architectural shift away running on a desktop computer can be integrated into the
from the classical single-micro-controller robot to a distributed, loop [6].
intrinsically modular, design. The resulting ability to design Another essential feature for the swarmanoid is communica-
and test components in isolation increases component quality tion between different robotic platforms. We have designed and
and allows for parallel development of different components. implemented a common sensing and communication system for
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 4

all the robots, based on a combination of infra-red and radio

communication. This system provides relative localisation and
structured communication signals. The system—referred to as
the range and bearing communication system—was inspired
by similar devices developed by Pugh and collaborators [7]
and by Gutiérrez and collaborators [8]. These previous devices,
however, presented severe limitations in the range and precision
of the communication system. We therefore took the decision to
design a novel integrated device. Our new device allows relative
localisation (from 10 cm up to 5 m for the foot-bots and hand-
bots and up to 12 m for the eye-bots), data communication at a
relatively high rate, and full-3D operation, all interference-free.
Our system uses a combination of new techniques to optimise
the way a range measurement is attained and how it transmits
the data. To obtain a measurement with an increased dynamic
range we use a four stage cascaded amplifier. Each of the
four stages is designed to output a voltage corresponding to
a complementary region of the maximum range. To optimise
the speed of a range measurement, we removed the data from
the infrared signal and instead transmit it over a 2.4 GHz
transceiver, which is also used to synchronise each range
and bearing system by implementing a simple turn taking
algorithm [9]. Fig. 3. The foot-bot robotic platform. The foot-bot has a differential drive
system that uses a combination of tracks and wheels to provide mobility
on rough terrain. Two of the foot-bots in this figure have illuminated their
B. Foot-bot LED communication ring. These RGB coloured signals are detectable by the
onboard cameras of other foot-bots.
The foot-bot (Figure 3) is an autonomous robot that improves
over the s-bot platform, previously developed within the Swarm-
bots project [10], [11], [12]. The robot is conceptually modular proprioceptive orientation measurement in all-terrain conditions,
at all levels: mechanics, electronics and software. Mechanical the foot-bot base includes 3-axis accelerometers and gyroscopes.
modularity is achieved by stacking modules on top of one All functionality contained in the base module is managed by
another, following well-defined specifications. The modularity three local dsPIC micro-controllers.
of the electronics is achieved by partitioning the required The gripper module is stacked above the base module and
functionality of each module to make them as independent as provides self-assembling abilities between the foot-bot and
possible. Each module is provided with its own local processing other foot-bots or hand-bots. Self-assembly is achieved through
power, therefore supporting the distributed architecture based a docking ring and a gripping mechanism with complementary
on ASEBA. The different modules share battery power, some shapes. The shape of the docking ring physically guides the
common control signals (e.g., power enable or reset), and the gripper into place, thus providing passive vertical alignment.
communications buses (CAN and I 2 C). The entire gripper module can be rotated around the foot-bot,
The foot-bot is 28 cm high and has a diameter of 13 cm. thus providing active horizontal positioning. A 2D force sensor
It is powered by a 3.7 V, 10 Ah Lithium-Polymer battery allows the foot-bot to measure the effort applied on the docking
contained in the base module, which also houses an “hot-swap” ring. This traction sensor is placed between the main structure
mechanism that allows battery exchange without switching off of the foot-bot body and the docking ring. Additionally, the
the robot. This capability is provided by a super-capacitor which module contains RGB LEDs enclosed inside the docking
maintains the power supply of the robot for 10 s during battery ring, which can be used for colour based communications
exchange. The foot-bot has differential drive motion control, with other foot-bots and hand-bots. The range and bearing
composed of two 2 W motors, each associated to a rubber module contains the local sensing and communication device
track and a wheel (referred to as “treels”). The maximum common to all the robots of the swarmanoid. It is very simple
speed of the foot-bot is 30 cm/s. The base of the foot-bot mechanically, but has complex analog electronics. The distance
includes infrared sensors, some acting as virtual bumpers and scanner module is based on 4 infrared distance sensors mounted
others as ground detectors. These sensors have a range of some on a rotating platform. We coupled two sensors of different
centimetres and are distributed around the robot on the main ranges ([40,300] mm and [200,1500] mm) to cover both short
printed circuit: 24 are outward-facing for obstacle detection, and long distances. The platform rotates continuously to make
8 are downward-facing for ground detection. Additionally, 4 360◦ scans. To minimise the wear and maximise the life time
contact ground sensors are placed under the lowest part of of the scanner, the fixed part transfers energy by induction to
the robot, in-between the treels. The base of the foot-bot also the rotating part, and the rotating and fixed parts of the module
contains an RFID reader and writer with an antenna situated exchange data using infrared light. Finally, the upper module
on the bottom of the robot, close to the ground. To allow for includes the cameras, a LED beacon, the i.MX31 ARM 11
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 5

processor and its peripherals such as WiFi board and flash card
reader. Two cameras are available: a top/front camera and an
omnidirectional camera.
Building on previous experience, the foot-bot design solves
many issues that we experienced in previous work with
the s-bot. The foot-bot is a much more stable platform. Its
slightly increased size (in comparison with the s-bot) and
modular design together allowed us to develop stronger and
higher quality components. The autonomy of the foot-bot is
improved thanks to new battery technology and to the hot-
swap mechanism, which enables longer experiments that are
not limited by battery life-time. The novel modular design
ensures flexibility of the system, which can be extended simply
by adding new components. For instance, new sensor modules
can be easily plugged in without the need to redesign the entire
robot or parts of it. In summary, the foot-bot is an excellent tool
for swarm robotics experimentation, as it features enhanced Fig. 4. Three hand-bots assembled together. The hand-bot is an autonomous
robot capable of climbing vertical structures and manipulating objects. The
autonomy, short and long range perception, robot-robot and grasping hands enable basic manipulation abilities, as well as the possibility
robot-environment interaction, self-assembling abilities and to physically connect to other hand-bots forming large structures.
a rich set of devices for sensing and communication. These
features are not currently found in any other collective robot
platform (among others, see [13], [14], [15], [16], [17], [18], attaching to ferromagnetic ceilings, a motor to switch the
[19], [20]). magnetic field and cancel the attachment force, a processor
controlling the system, an IR receiver to get commands from
the hand-bot and super-capacitors to store the energy to power
C. Hand-bot
supply the system. The whole system requires 1.4 mA for
The hand-bot has no autonomous mobility on the ground, but standby power supply and can survive powered on for 35
is able to climb standard office furniture, grasp small objects minutes. When switched on, the magnet can provide a vertical
such as books or letters, and bring such objects to the ground. force of 140 N [22]. The launcher mechanism has been designed
For the swarmanoid to transport an object, the hand-bot can with reliability in mind, both in launching and in retrieving the
grasp the object while itself being transported by the foot-bots. rope. The upper part of the launcher contains RGB LEDs that
The hand-bot can thus interact physically with other robots of can be used for signalling between robots. Two fan propellers
the swarmanoid. attached to the launcher provide the hand-bot with orientation
In the literature, it is possible to find a large number of climb- and limited position control while suspended to the rope.
ing robots, which rely on different techniques to implement the The main body of the hand-bot protects the launcher
climbing mechanism. For a recently published overview of the mechanisms and hosts a number of devices. In the front
existing climbing systems, see [21]. In designing the hand-bot, part, a high resolution camera looks forward towards the area
we considered magnetic attachment systems, grasping hands, accessible by the grasping hands. The battery—identical to that
suction pads, dry adhesion mechanisms and mechanisms based of the foot-bot—is housed within the main body, as is the range
on some external aid, such as ropes or poles. Given the currently and bearing system and the docking ring. The range and bearing
available technologies, the solution we settled on for the hand- and the ring are identical in functionality to those of the foot-bot,
bot is a combination of several approaches, namely grasping but have been modified in order to fit the shape of the hand-bot.
hands seconded by a climbing assistance device based on a rope Around the main body, the docking ring allows connections
launcher and a magnetic attachment system. The rope ensures from foot-bots. The ring contains 12 RGB LEDs for visual
vertical movement without the need of strong attachment to the signalling. Finally, the hand-bot features two identical arms,
walls. The rope can be launched from the hand-bot to attach which provide climbing and manipulation abilities. The arms
to the desired position on the ceiling. For multiple launches, are parallelogram-based structures that ensure the alignment
the hand-bot can actively detach and retrieve the rope, before of the two grippers with the body. The two arms are mounted
recharging the system in preparation for the next launch. The symmetrically on the central rotating system—the head—and
grasping hands ensure connections to vertical structures and provide one independent and one coupled degree of freedom
the ability to manipulate objects (see Figure 4). The hand-bot to each gripper, for a total of three degrees of freedom. Each
is 29 cm high, 41 cm wide in its widest configuration (with its grasping hand contains an embedded low resolution colour
arms fully retracted) and 47 cm long in its longest configuration camera (VGA) and 12 distance sensors, which can be used
(with its arms fully extended). in conjunction to locate and grasp objects in the environment.
The rope launcher and the magnetic system modules are The gripper was designed to support the weight of the robot
the most challenging parts of the robot design because of when the arms are in a vertical position. This implies a high
the complexity of the device and the robustness required by grasping force of 25 N. The gripper can also rotate with a load
its operation. The attachment system includes the magnet for of 2 N (e.g., the weight of a book). The gripper is designed
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 6

also to allow a firm connection to the arms of other hand-bots,

which in this way may form a physically connected structure,
as shown in Figure 4. By launching their attachment system to
the ceiling, assembled hand-bots can climb and control their
position in the 3D space (for more details, see [23]).
In summary, the hand-bot is designed as a compact robot
dedicated to climbing and manipulation scenarios. At the
electronic level, the robot has an architecture identical to
the foot-bot and shares most of the basic components. It is
similarly modular and also supports the ASEBA architecture.
Many components are shared with the foot-bot and eye-bot,
such as the i.MX31 processor board, the motor controllers, the
range and bearing system and the battery.

D. Eye-bot
The eye-bot is an autonomous flying robot designed to
operate in indoor environments (see Figure 5). The eye-bots
work in synergy with the rest of the swarmanoid: they provide
an aerial view to detect the objects of interest and to direct
the actions of other robot types. The size of an eye-bot has
been optimised to obtain a small enough platform capable
of flying in a large room without interfering with other (a)
platforms, and capable of flying in narrow corridors to explore
the environment. Innovative methods have been employed to
dramatically increase mission endurance: the eye-bot features
a ceiling attachment system that enables an energy saving
operation mode in which the eye-bot can power down its flight
systems, but continue to scan the environment and communicate
with the rest of the swarmanoid.
The eye-bot has been designed around an advanced quadrotor
structure, which allowed us to reduce the size of the robot
without sacrificing payload capability or flight endurance.
Recent advances have permitted the stable control of small (b)
hover-capable robots like quadrotors [24]. However, although Fig. 5. The eye-bot platform. (a) The ceiling attachment system and the
altitude stability is feasible, hovering robots usually suffer from 24 × 6.5 × 6.5 cm rectangular structure housing the batteries and main PCBs.
drift. Platform drift is an unavoidable result of imbalances in (b) The four contra-rotating coaxial rotors, the circular 3D range and bearing
communication system, and the pan-tilt camera with the laser pointer.
the rotor blades, differing air-flow over the airframe, turbulence
from down-wash or external forces such as wind. This drift
is commonly compensated for with absolute positioning. In
outdoor systems, absolute positioning usually relies on GPS. many limitations and only function within certain environments.
Absolute positioning indoors has been implemented using In contrast, the eye-bots are collectively capable of autonomous
colour vision cameras [25] or infrared 3D motion tracking flight without any of these limitations. Flying eye-bots can
cameras, e.g., the Vicon system [26]. Such tracking systems manoeuvre using sensory information from other static eye-
provide high-accuracy measurements of position and altitude bots, communicated over the on-board range and bearing
at fast refresh rates (1-5 mm at 200 Hz), allowing the control communication system. By having at least one eye-bot attached
of a small aircraft in highly dynamic manoeuvres such as to the ceiling that provides a static reference point, it is possible
multi-flip trajectories [26]. However, this approach requires to control the unknown egomotions and the platform drift. A
an environment that has been tailored in advance with the cooperating network of eye-bots attached to the ceiling [30] thus
installation of the relevant sensors, which may not always be enables indoor navigation whilst avoiding the use of absolute
feasible. Common approaches to autonomous flight with on- positioning systems such as GPS, the pre-installation of 3D
board sensors exploit either laser scanners or visual processing tracking cameras, illumination dependent visual processing or
[27], [28]. Laser scanners are heavy and computationally computationally expensive laser scan-line matching.
expensive, while vision-based approaches are highly dependent The eye-bot uses a quadrotor-like propulsion configuration
on the available ambient illumination, which may be insufficient but with a 4x2 co-axial rotor system (see Figure 5(b)). Each
or unpredictable in many situations. Similar problems affect rotor system consists of a co-axial counter-rotating brushless
optic-flow approaches which require significant environment motor (Himax Outrunner HC2805-1430) which provides 500 g
texture and contrast [29]. In summary, previous approaches have thrust at 9 V (750 g at 12 V). This gives a total platform thrust of
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 7

Controller In traditional simulator designs, such as those of Webots [31],

USARSim [32] and Gazebo [33], accuracy is the main driver,
Control Interface
at the cost of limited scalability. Simulators designed for
Sensors Actuators scalability, such as Stage [34], are focused on very specific
application scenarios, thus lacking flexibility. To achieve both
scalability and flexibility, in the design of ARGoS we made a
Entities number of innovative choices.
ARGoS’ architecture is depicted in Figure 6. Its core, the
Simulated 3D Space
simulated space, contains all the data about the current state
of the simulation. Such data is organized into sets of entities
of different types. Each entity type stores a certain aspect of
Physics Engines Visualizations the simulation. For instance, positional entities contain the
position and orientation of each object in the space. Entities
Fig. 6. The architecture of the ARGoS simulator. are also organized into hierarchies. For example, the embodied
entity is an extension of the positional entity that includes a
bounding box. Robots are represented as composable entities,
at least 2000 g, sufficient to lift the payload for the advanced that is, entities that can contain other entities. Each individual
sensory-motor systems. The main body has a carbon fibre robot feature is stored into dedicated entity types. For instance,
structure, and houses the batteries and main printed circuit each robot possesses an embodied entity and a controllable
boards (PCBs) such as the flight computer and i.MX31 ARM entity, that stores a pointer to that robot’s sensors, actuators
11 processor. Attached to the bottom of the body structure is and control code.
the propulsion system consisting of 4 carbon fibre arms that Organizing data in the simulated space in this way provides
support the motors, the rotary systems and the range and bearing both scalability and flexibility. Scalability is achieved by
module. On top of the eye-bot resides the ceiling attachment organizing entities into type-specific indexes, optimized for
mechanism. Finally, the eye-bot has 4 carbon fibre legs for speed. For instance, all positional entities are organized into
support. These legs also protect the rotors and the delicate pan- space hashes, a simple and state-of-art technique to store and
tilt camera system. In total, the carbon fibre structure weighs retrieve spatial data. Flexibility is ensured because entities are
only 270 g. The outer diameter is 50 cm and the total height implemented as modules. In addition to the entities offered
including the legs and ceiling attachment is 54 cm. natively by ARGoS, the user can add custom modules, thus
As mentioned above, the eye-bot is reliant on the range and enriching ARGoS’ capabilities with novel features.
bearing communication device. This communication system Analogously, the code accessing the simulated space is
allows an eye-bot to communicate with other eye-bots, to organized into several modules. Each individual module can
coordinate movements in 3D and to facilitate controlled flight be overridden by the user whenever necessary, thus ensuring
without platform drift. The system is fully compatible with a high level of flexibility. The modules are implemented as
the similar devices developed for the foot-bot and the hand- plug-ins that are loaded at run-time.
bot, and permits bi-directional communication between the Controllers are modules that contain control code developed
different robotic platforms. The system mounted on the eye- by the user. Controllers interact with a robot’s devices through
bots provides the range and bearing of robots within 12 m, as an API called the common interface. The common interface
well as low-bandwidth local communication. API is an abstraction layer that can make underlying calls to
Inter-robot communication can also take place via colour- either a simulated or a real-world robot. In this way, controllers
based visual signals. An array of RGB LEDs around the can be seamlessly ported from simulation to reality and back,
perimeter of the eye-bot can be illuminated in different colour making behaviour development and its experimental validation
patterns. To view the colour LED rings of other robots and more efficient.
to detect target objects of interest, the eye-bots are equipped Sensors and actuators are modules that implement the
with a high-resolution colour CMOS camera mounted on a common interface API. Sensors read from the simulated space
2-axis pan-tilt mechanism. This allows the eye-bot to have high and actuators write to it. The optimized entity indexes ensure
resolution imaging in the volume of space beneath the eye-bot. fast data access. For each sensor/actuator type, multiple imple-
The same pan-tilt mechanism additionally holds a 5 mW Class mentations are possible, corresponding to models that differ in
IIIA laser pointer. This laser can be pointed in any direction computational cost, accuracy and realism. In addition, sensors
beneath the eye-bot. and actuators are tightly coupled with robot component entities.
For instance, the foot-bot wheel actuator writes into the wheeled
equipped entity component of the foot-bot. Such coupling
E. Simulation
greatly enhances code reuse. New robots can be inserted by
ARGoS is a novel simulator designed to simulate the combining existing entities, and the sensors/actuators depending
swarmanoid robots and to enable fast prototyping and testing on them work without modification.
of robot controllers. The main features of ARGoS are high Visualizations read the simulated space to output a representa-
scalability for increasing numbers of robots and high flexibility tion of it. Currently, ARGoS offers three types of visualization:
to allow users to add functionality. (i) an interactive GUI based on Qt and OpenGL, (ii) a high
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 8

(a) (b)

Fig. 7. Screen-shots from different visualizations. (a) Qt-OpenGL. (b) POV-Ray.

quality off-line 3D renderer based on POV-Ray, and (iii) a space (such as an office building). The swarmanoid robots
textual renderer designed to interact with data analysis and are assumed to start from a single deployment area. The
plotting software such as Matlab and GNUPlot. Figure 7 shows swarmanoid must first find the shelves containing relevant
some of the visualization possibilities of ARGoS. objects, and then transport the objects from the shelves back
One of the most distinctive features of ARGoS is that the to the deployment area.
simulated space and the physics engine are separate concepts. The swarmanoid search and retrieval behaviour we developed
The link between them is the embodied entity, which is stored is given in Figure 8. Eye-bots collectively explore the environ-
in the simulated space and updated by a physics engine. In ment and search for the target location. They slowly build a
ARGoS, multiple physics engines can be used simultaneously. wireless network that spans the environment by connecting to
In practice, this is obtained by assigning sets of embodied the ceiling. Each new flying eye-bot that joins the search is
entities to different physics engines. The assignment can be guided to the edge of the network by the eye-bots already in
done in two complementary ways: (i) manually, by binding place. Having reached the edge of the network, the searching
directly an entity to an engine, or (ii) automatically, by eye-bot continues flying, thus exploring new terrain. The eye-
assigning a portion of space to the physics engine, so that every bot will, however, stop flying and attach to the ceiling when
entity entering that portion is updated by the corresponding at the limit of its communication range with the rest of the
engine. Physics engines are a further type of module. Currently,
three physics engines are available: (i) a 3D dynamics engine
based on the ODE library, (ii) a 2D dynamics engine based on
the Chipmunk library, and (iii) a custom-made 2D kinematic
Foot-hand-bot Eye-bot
engine. self-assembly swarm search
To further enhance scalability, the architecture of ARGoS Gives
Gives
control communicates the
is multi-threaded. The simulation loop is designed in such control
to to
communicates the shelf distance and
shelf and nest direction
a way that race conditions are avoided and that CPU usage Foot-hand-bot
distance
direction
and

Foot-bot chain
is optimized. The parallelization of the calculations of sen- navigation

sors/actuators and of the physics engines provides high levels Gives

control
to
of scalability. Results reported in [35] show that ARGoS can
communicates the position
simulate 10 000 simple robots 40% faster than real time. Foot-hand-bot of the book

2 shelf aligment
ARGoS has been released as open source software and
Gives Gives
currently runs on Linux and Mac OS X. control control
to to

Hand-bot object
III. S WARMANOID I N ACTION retrieval Gives
control
to
A. Search and Retrieval: Behavioural Control
To demonstrate the potential of the swarmanoid concept,
Retract
we developed an integrated search and retrieval behaviour. Foot-bot Hand-bot
The search and retrieval behaviour is designed to allow the assembly

swarmanoid to retrieve objects in a complex 3D environment.

Objects are placed on one or more shelves in a human habitable
Fig. 8. The general schema of the scenario behavioural components and their
2 http://iridia.ulb.ac.be/argos/ interactions
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 9

network. The network remains connected using the range and

bearing communication system [30].
To free up potentially scarce eye-bot resources, foot-bots
incrementally form a complementary wireless network on the
ground that follows the eye-bot network topology, but extends
only in the most promising search directions identified by the
eye-bots. The eye-bot network and the foot-bot network can
pass range and bearing messages between each other, and thus
act as an integrated heterogeneous network. As the slower foot-
bot network catches up with the eye-bot network, eye-bots are
freed up for further exploration. Thus the eye-bots provide a
fast and systematic exploration of the environment, while foot-
bots provide longer term storage of exploration information on
the ground. (a)
Whenever an exploring eye-bot finds a shelf containing
objects, it communicates knowledge of its discovery back
to the nest through the heterogeneous network of eye-bots
and foot-bots. The swarmanoid now needs hand-bots at the
shelf location to retrieve the objects. In the deployment area,
foot-bots thus assemble to hand-bots and start collectively
transporting them to the shelf [36]. We refer to the composite
entity formed by the foot-bots assembled to a hand-bot as a
foot-hand-bot (see Figure 1(a)). Guided by the heterogeneous
robotic network of eye-bots and foot-bots, the foot-hand-bots
can navigate through the environment following the shortest
path from the nest to the shelf. When the foot-hand-bot arrives
at a shelf location, the eye-bot that found the shelf conveys
(b)
information about the 3D location of a particular object on
the shelf to the foot-hand-bot. The information tells the foot- Fig. 9. Drawing of the test arena for the swarmanoid demonstration. (a)
hand-bot’s constituent hand-bot where it should climb and to Parallel projection. (b) Floor plan.
what height. To allow the hand-bot to climb, the foot-hand-bot
disassembles, and the constituent foot-bots retreat. These foot-
bots wait for the hand-bot at the foot of the shelf and act as
markers to subsequent foot-hand-bots letting them know not
to approach the shelf at that location. The hand-bot retrieves
the book, and descends from the shelf. The foot-hand-bot then
re-assembles and follows the heterogeneous chain back to the
nest.

B. Search and Retrieval: Real World Demonstration

We demonstrated our integrated search and retrieval be-
haviour in a real-world demonstration. Our demonstration
involved a real-world instantiation of the generic search and
retrieval task, in an environment containing a single shelf
and book. The arena we used can be seen in Figure 9. We
successfully demonstrated that a swarmanoid with no a priori
knowledge of the environment was able to find the shelf and
retrieve the book. This scenario integrated various swarmanoid
abilities, ranging from task allocation to collective search, from
self-assembly to cooperative transport, from object retrieval to
cooperative navigation in complex environments. Fig. 10. A snapshot of the video demonstrating the swarmanoid involved
Figure 10 shows a snapshot from a video of a successful in the object retrieval experimental scenario. The video is shot from four
experiment. The video shown in Figure 10 is available as different cameras simultaneously, giving full coverage of a single experiment.
This video is available in the supplementary electronic material. (Top Left)
supplementary electronic material for this paper. A separate View from deployment area towards doorway. (Top Right) View of doorway.
video addressed to the general public, edited together from (Bottom Left) View from doorway towards shelf. (Bottom Right) View of
different experiments, has been submitted to the AAAI 2011 shelf.
Video Competition and can be viewed at http://iridia.ulb.ac.be/
swarmanoid-the-movie.
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 10

C. Further Experimentation R EFERENCES

Our success in realising the experimental search and rescue [1] E. Bonabeau, M. Dorigo, and G. Theraulaz, Swarm Intelligence: From
scenario demonstrates the viability of the swarmanoid concept, Natural to Artificial Systems. Oxford University Press, New York, NY,
and gives a concrete example of how heterogeneous swarms 1999.
[2] M. Dorigo and E. Şahin, “Swarm robotics — special issue editorial,”
can solve complex problems. The abilities of the swarmanoid Autonomous Robots, vol. 17, no. 2–3, pp. 111–113, 2004.
are not, however, limited to the scenario we presented above. [3] V. Trianni, Evolutionary Swarm Robotics. Evolving Self-Organising Be-
The swarmanoid can in principle carry out a wide range of haviours in Groups of Autonomous Robots, ser. Studies in Computational
Intelligence. Springer Verlag, Berlin, Germany, 2008, vol. 108.
tasks that can benefit from parallel operation and leverage the [4] P. Grassé, “La reconstruction du nid et les coordinations interindi-
different abilities of the three robot types. viduelles chez Bellicositermes natalensis et Cubitermes sp. La théorie
Within the swarmanoid framework, we have carried a number de la stigmergie: Essai d’interprétation du comportement des termites
constructeurs,” Insectes Sociaux, vol. 6, pp. 41–81, 1959.
of experiments, both in simulation and with physical robots. The [5] R. Groß and M. Dorigo, “Self-assembly at the macroscopic scale,”
development of control algorithms for the swarmanoid followed Proceedings of the IEEE, vol. 96, no. 9, pp. 1490–1508, sep 2008.
multiple research lines. On the one hand, behaviour based [6] S. Magnenat, P. Rétornaz, M. Bonani, V. Longchamp, and F. Mondada,
“ASEBA: A modular architecture for event-based control of complex
approaches have been employed for tasks such as recruitment, robots,” IEEE/ASME Transactions on Mechatronics, vol. 16, no. 2, pp.
collective transport, collective exploration and so forth [37], 321–329, 2011.
[36], [38]. On the other hand, evolutionary robotics techniques [7] J. Pugh and A. Martinoli, “Relative localization and communication
have been used to synthesise efficient neural network controllers module for small-scale multi-robot systems,” in Proceedings of the 2006
IEEE International Conference on Robotics and Automation (ICRA 2006),
for behavioural synchronisation and for path formation between 2006, pp. 188–193.
two target areas [39], [40]. All these studies demonstrate the [8] Á. Gutiérrez, A. Campo, M. Dorigo, D. Amor Martı́n, L. Magdalena,
potential of heterogeneous robotic swarms, and point to a new and F. Monasterio-Huelin Maciá, “An open localisation and local
commnunication embodied sensor,” Sensors, vol. 8, no. 11, pp. 7545–
way of tackling complex application scenarios in the real world. 7563, 2008.
[9] J. Roberts, T. Stirling, J.-C. Zufferey, and D. Floreano, “2.5D infrared
range and bearing system for collective robotics,” in 2009 IEEE/RSJ
IV. S UMMARY AND C ONCLUSIONS International Conference on Intelligent Robots and Systems (IROS 2009),
2009, pp. 3659–3664.
Advancements of the state of the art in swarm robotics [10] F. Mondada, G. Pettinaro, A. Guignard, I. Kwee, D. Floreano, J.-L.
can be pursued by relying on heterogeneous swarm systems Deneubourg, S. Nolfi, L. Gambardella, and M. Dorigo, “SWARM-BOT:
composed of a large number of robots presenting behavioural A new distributed robotic concept,” Autonomous Robots, vol. 17, no.
2–3, pp. 193–221, 2004.
and/or physical heterogeneities. To this end, it is necessary [11] M. Dorigo, “SWARM-BOT: An experiment in swarm robotics,” in
to develop tools and methodologies that enable the use of Proceedings of the 2005 IEEE Swarm Intelligence Symposium (SIS-05),
such heterogeneous systems. We identified relevant issues and P. Arabshahi and A. Martinoli, Eds., 2005, pp. 192–200.
[12] M. Dorigo, E. Tuci, V. Trianni, R. Groß, S. Nouyan, C. Ampatzis,
challenges, in particular highlighting the difficulty of delivering T. Labella, R. O’Grady, M. Bonani, and F. Mondada, “SWARM-BOT:
the tightly integrated robotic hardware necessary to enable Design and implementation of colonies of self-assembling robots,” in
physical and behavioural interaction between different robot Computational Intelligence: Principles and Practice, G. Yen and D. Fogel,
Eds. IEEE Computational Intelligence Society, New York, NY, 2006,
types. ch. 6, pp. 103–135.
We presented the swarmanoid as a new robotic concept in [13] S. Kernbach, , E. Meister, F. Schlachter, K. Jebens, M. Szymanski,
heterogeneous swarm robotics. The hardware and the software J. Liedke, D. Laneri, L. Winkler, T. Schmickl, R. Thenius, P. Corradi, and
L. Ricotti, “Symbiotic robot organisms: REPLICATOR and SYMBRION
of the swarmanoid robots leveraged common technologies to projects,” in Proceedings of the Performance Metrics for Intelligent
ensure seamless integration of the different platforms. The Systems Workshop (PerMIS), 2008, pp. 62–69.
resulting compatibility of different robot types enabled us [14] K. Konolige, D. Fox, C. Ortiz, A. Agno, M. Eriksen, B. Limketkai, J. Ko,
B. Morisset, D. Schulz, B. Stewart, and R. Vincent, “Centibots: Very
to explore different coordination mechanisms and strategies large scale distributed robotic teams,” in Experimental Robotics IX, ser.
in a heterogeneous swarm. The experimental scenario we Springer Tracts in Advanced Robotics, M. H. Ang and O. Khatib, Eds.
defined demonstrates the suitability of the swarmanoid robotic Springer Verlag, Berlin, Germany, 2006, vol. 21, pp. 131–140.
[15] A. Turgut, F. Gokce, H. Celikkanat, L. Bayindir, and E. Şahin., “Kobot:
concept for tackling complex problems in 3D human-made A mobile robot designed specifically for swarm robotics research,” Dept.
environments. Future work will use the swarmanoid robotic of Computer Engineering, Middle East Technical University, Ankara,
platforms to develop a rigorous methodological approach Turkey, Tech. Rep. METU-CENG-TR-2008-01, 2008.
[16] R. Rusu, G. Lazea, R. Robotin, and C. Marcu, “Towards open archi-
for the design of behaviours for swarm robotics systems, tectures for mobile robots: ZeeRO,” in Proceedings of the 2006 IEEE
especially focusing on hierarchical, heterogeneous control and International Conference on Automation, Quality and Testing, Robotics
communication. (AQTR 2006), 2006, pp. 260–265.
[17] V. Zykov, P. William, N. Lassabe, and H. Lipson, “Molecubes extended:
Diversifying capabilities of open-source modular robotics,” in IROS-2008
ACKNOWLEDGMENTS Self-Reconfigurable Robotics Workshop, 2008.
[18] P. Rybski, A. Larson, H. Veeraraghavan, M. Anderson, and M. Gini,
This work was supported by the Swarmanoid project, funded “Performance evaluation of a multi-robot search & retrieval system:
by the Future and Emerging Technologies programme (IST- Experiences with MinDART,” Journal of Intelligent and Robotic Systems,
vol. 52, no. 3, pp. 363–387, 2008.
FET) of the European Commission under grant IST-022888. [19] K. Yoo, Y. Lee, H. Choi, B. Yoo, and D. Kim, “Swarm robotics: Self
The information provided is the sole responsibility of the assembly, physical configuration, and its control,” in SICE-ICASE, 2006.
authors and does not reflect the European Commission’s International Joint Conference, 2006, pp. 4276–4279.
[20] H. Zhang, W. Wang, Z. Deng, G. Zong, and J. Zhang, “A novel
opinion. The European Commission is not responsible for any reconfigurable robot for urban search and rescue,” International Journal
use that might be made of data appearing in this publication. of Advanced Robotic Systems, vol. 3, no. 4, pp. 359–366, 2006.
IEEE ROBOTICS & AUTOMATION MAGAZINE, VOL. X, NO. X, MONTH 20XX 11

[21] B. Chu, K. Jung, C.-S. Han, and D. Hong, “A survey of climbing

robots: Locomotion and adhesion,” International Journal of Precision
Engineering and Manufacturing, vol. 11, pp. 633–647, 2010.
[22] F. Rochat, P. Schoeneich, M. Bonani, S. Magnenat, F. Mondada,
H. Bleuler, and H. Christoph, “Design of magnetic switchable device
(MSD) and applications in climbing robot,” in Emerging Trends in Mobile
Robotics, H. Fujimoto, M. O. Tokhi, H. Mochiyama, and G. S. Virk,
Eds. World Scientific, 2010, pp. 375–382.
[23] M. Bonani, P. Rétornaz, S. Magnenat, H. Bleuler, and F. Mondada,
“Physical interactions in swarm robotics: the hand-bot case study,”
in Proceedings of the 10th International Symposium on Distributed
Autonomous Robotic Systems (DARS 2010), 2010.
[24] E. Altug, J. Ostrowski, and C. Taylor, “Control of a quadrotor helicopter
using dual camera visual feedback,” The International Journal of Robotics
Research, vol. 24, no. 5, pp. 329–341, 2005.
[25] G. Hoffmann, S. Waslander, and C. Tomlin, “Quadrotor helicopter
trajectory tracking control,” in Proceedings of the AIAA Guidance,
Navigation, and Control Conference. AIAA, 2008, pp. 1–14.
[26] S. Lupashin, A. Schöllig, M. Sherback, and R. D’Andrea, “A simple
learning strategy for high-speed quadrocopter multi-flips,” in Proceedings
of the 2010 IEEE International Conference on Robotics and Automation
(ICRA 2010), 2010, pp. 1642–1648.
[27] F. Kendoul, K. Nonami, I. Fantoni, and R. Lozano, “An adaptive vision-
based autopilot for mini flying machines guidance, navigation and control,”
Autonomous Robots, vol. 27, no. 3, pp. 165–188, 2009.
[28] N. Guenard, T. Hamel, and R. Mahony, “A practical visual servo control
for a unmanned aerial vehicle,” IEEE Transactions on Robotics, vol. 24,
no. 2, pp. 331–341, 2008.
[29] A. Beyeler, J.-C. Zufferey, and D. Floreano, “3D vision based navigation
for indoor microflyers,” in Proceedings of the 2007 IEEE International
Conference on Robotics and Automation (ICRA 2007), 2007, pp. 1336–
1341.
[30] T. Stirling, S. Wischmann, and D. Floreano, “Energy-efficient indoor
search by swarms of simulated flying robots without global information,”
Swarm Intelligence, vol. 4, no. 2, pp. 117–143, 2010.
[31] O. Michel, “Cyberbotics Ltd. – Webots: Professional mobile robot
simulation,” International Journal of Advanced Robotic Systems, vol. 1,
no. 1, pp. 39–42, March 2004.
[32] S. Carpin, M. Lewis, J. Wang, S. Balakirsky, and C. Scrapper, “USARSim:
a robot simulator for research and education,” in Proceedings of the
IEEE Conference on Robotics and Automation (ICRA). Piscataway, NJ:
IEEE Press, 2007, pp. 1400–1405.
[33] N. Koenig and A. Howard, “Design and use paradigms for Gazebo,
an open-source multi-robot simulator,” in Proceedings of the IEEE/RSJ
International Conference on Intelligent Robots and Systems (IROS).
Piscataway, NJ: IEEE Press, 2004, pp. 2149–2154.
[34] R. Vaughan, “Massively multi-robot simulation in Stage,” Swarm
Intelligence, vol. 2, no. 2, pp. 189–208, 2008.
[35] C. Pinciroli, V. Trianni, R. O’Grady, G. Pini, A. Brutschy, M. Bram-
billa, N. Mathews, E. Ferrante, G. Di Caro, F. Ducatelle, T. Stirling,
A. Gutiérrez, L. M. Gambardella, and M. Dorigo, “ARGoS: a modular,
multi-engine simulator for heterogeneous swarm robotics,” in Proceedings
of the IEEE/RSJ International Conference on Intelligent Robots and
Systems (IROS 2011), 2011, in press.
[36] E. Ferrante, M. Brambilla, M. Birattari, and M. Dorigo, “Socially-
mediated negotiation for obstacle avoidance in collective transport,”
in Proceedings of the 10th International Symposium on Distributed
Autonomous Robotic Systems (DARS 2010), 2010.
[37] C. Pinciroli, R. O’Grady, A. Christensen, and M. Dorigo, “Self-organised
recruitment in a heterogeneous swarm,” in Proceedings of the 2009
International Conference on Advanced Robotics (ICAR 2009), E. Prassel
et al., Eds., 2009, pp. 1–8.
[38] F. Ducatelle, G. Di Caro, C. Pinciroli, and L. Gambardella, “Self-
organized cooperation between robotic swarms,” Swarm Intelligence,
vol. 5, no. 2, pp. 73–96, 2011.
[39] V. Trianni and S. Nolfi, “Self-organising sync in a robotic swarm. A dy-
namical system view,” IEEE Transactions on Evolutionary Computation,
vol. 13, no. 4, pp. 722–741, 2009.
[40] V. Sperati, V. Trianni, and S. Nolfi, “Self-organised path formation in a
swarm of robots,” Swarm Intelligence, vol. 5, no. 2, pp. 97–119, 2011.