Aerospace and Electronic Systems - March 2019 - 40

Feature Article:

DOI. No. 10.1109/MAES.2019.2905922

ARIES: An Autonomous Controller For Multirobot
Cooperation
~
,
Fernando Ropero, Pablo Munoz,
Marıa D. R-Moreno, University of Alcala
de Henares, Spain
Alcala

INTRODUCTION
Over the last decades, the technological growth in
advanced autonomous robots, such as unmanned aerial
vehicles (UAVs) or autonomous underwater vehicles
(AUVs), has been enhancing novel solutions where groups
of robots work together to perform complex tasks. Space
agencies have realized that robotic cooperation represents
a feasible path for the space exploration because it can
provide new capabilities to improve both the quality and
quantity of the scientific return. This is reflected in the
Mars 2020 mission, in which NASA plans to launch the
Mars Helicopter Scout (MHS) to demonstrate the aerial
technology required for the Mars exploration and to establish the path for cooperative exploration with a rover on
Mars. The objective is that the MHS will recognize the surroundings of the rover to provide high-resolution aerial
images of the terrain for science (identifying potential targets) and operational (avoid driving through hazardous
zones) purposes. Nonetheless, the technology required for
such cooperation is still not mature enough; thus, the Mars
2020 mission will be still commanded by human operators.
To enable an autonomous coordinated mission involving multiple robots, there are several issues to solve. Meanwhile single-robot autonomy is a widely studied field, the
cooperation of multiple autonomous robots in the same
environments implies new problems and risks that shall be
taken into consideration. For instance, questions such as
collision avoidance, tasks optimization, and risks mitigation have to be integrated in the cooperation schema. In this
direction, defining cooperation policies to control and coordinate a multiple robot team is a tedious task due to the high

Authors' current address:
F. Ropero, P. Mun˜oz,
M. D. R-Moreno, Computer Engineering Department,
University of Alcala, 28871 Alcala´ de Henares, Spain.
E-mail: (fernando.ropero@uah.es).
Manuscript received September 30, 2018; revised
December 13, 2018. ready for publication January 23, 2019.
Review handled by Silvia ULLO and Peter Willett.
0885/8985/19/$26.00 ß 2019 IEEE
40

number of factors to assess. Moreover, in an environment
like Mars, there are extra constraints due to the limited
knowledge of the environment, such as wind speed or
atmospheric density, which cannot be easily modeled neither predicted, hardening the deployment of a robust system. This ignited our motivation and driven us to bring out
a research line to which the present work belongs to.
Autonomous cooperation represents a current research
topic devising autonomous controllers to implement cooperation paradigms for enabling the coordination among the
robotic platforms. Numerous works in autonomous controllers [a synthesis was written by Wooldridge, (2009)] present studies toward a technology to tackle challenges such
as the space exploration with a cooperative robot team.
Autonomous controllers or autonomous control architectures are widely known for achieving robot autonomy by
means of the sense-plan-act cycle. To do so, these systems
are usually implemented connecting different layers, each
one relying on a different technology than the previous one.
This is the case of the three tiers (3T) architectures [1], a
common scheme for implementing autonomous robotic
controllers which are, usually, built for specific robotic platforms and hardly reusable in different contexts than the initial one. Some architectures based on this scheme are: the
one produced by the Laboratory of Analysis and Architecture of Systems (LAAS) [2], the Coupled Layer Architecture for Robotic Autonomy (CLARAty) [3], or the Tripodal
Schematic Control Architecture [4]. An evolution of the 3T
scheme is the multiagent architecture, such as the intelligent distributed execution architecture [5] or the TeleoReactive EXecutive (T-REX) [6]. These architectures are
split up into agents where each agent can integrate a 3T
schema to control a subsystem of a single robotic platform
or a complete robot. Thus, it is possible to deploy several
agents for controlling multiple robotic platforms. Nevertheless, these schemas are typically executed in a centralized
framework in which a cooperative scenario might be tough
to deploy if flexibility, scalability, and fault-tolerant policies need to be considered as key factors.
This paper presents the Autonomous coopeRatIve
Execution System (ARIES), an autonomous controller

IEEE A&E SYSTEMS MAGAZINE

MARCH 2019

Aerospace and Electronic Systems - March 2019

Table of Contents for the Digital Edition of Aerospace and Electronic Systems - March 2019