MINISCUOLA WOA 2008

Palermo, 18 Novembre

From Software Agents to Physical Agents: an Overview of Issues in Autonomous Mobile Robot Programming

Corrado Santoro, Università di Catania

Autonomous Mobile Robots share many characteristics with software agents: they are situated in a well-know environment (which is physical/real for robots), they are goal-oriented, they have to achieve the goal in a complete autonomy basing their behaviour only on sensing the environment by means of proper sensors. In the same way as to multi-agent systems, multi-robot systems use the concepts of cooperative work and collaborative problem solving to achieve a complex goal that each robot alone would not be able to carry on. In this field, cooperative algorithms and approaches have been developed for some applications like load carrying, object/bomb discovery, environment monitoring, etc., often inspired from biological behaviours of colonies, like ants, termites, bird flocks, bees. Even if autonomous robots and multi-robot systems share such properties with the software agent world, they feature different problems from the software architecture point of view, that need to be faced using a proper approach.

On this basis, the talk will focus on such an aspect, stating the problems that have to be solved in autonomous mobile robot programming and presenting a robotic framework, characterized by a layered architecture able to clearly separate in different modules the various tasks concurring in implementing the overall control system of the robot.

The framework has been implemented using the Erlang language and has been employed in the robot made by the components of the team from the University of Catania which participated to Eurobot 2008 competition.

A tool for simulating the functioning of the robot will be also described, based on the ROSEN tool, a 3D robotic simulation engine. Thanks to ROSEN, the control system implemented using the framework can be tested by means of a simulation approach, before being deployed onto the real robot. A brief overview of how to use ROSEN  to simulate swarm robotic environments is also given.

The preliminary outline of the talk is the following:

1.      Robots and Agents: similarities and differences

2.      Issues in autonomous mobile robot programming

3.      The robotic framework

4.      ROSEN

5.      Glueing ROSEN with the robotic framework

6.      A step forward: using ROSEN to simulate swarm robotic environments.

Biography. See http://www.diit.unict.it/users/csanto/

How to solve relevant problems in bioinformatics using agents

Eloisa Vargiu, DIEE, Università di Cagliari

Bioinformatics is an emerging discipline that use information technology to organize, analyze, and distribute biological information in order to answer complex biological questions. Nowadays most of the software systems, tools, and applications in bioinformatics are very complex and require high flexibility. It is difficult to identify a common underlying technology since each specific research field imposes its own constraints on the corresponding solutions. Nevertheless, from a software engineering perspective, the agent technology appears to fit the requirements imposed by such research field and encompasses most of the bioinformatics systems, tools, and applications in a unifying framework.

This tutorial, after an introduction on bioinformatics, will focus on the peculiarities of a multiagent approach to bioinformatics applications. In particular, it will put into evidence the scenarios in which agents can be useful in such research field depicting some of the most interesting applications currently proposed in the literature.

Biography. Eloisa Vargiu obtained her Ph.D. in Electronic and Computer Engineering from the University of Cagliari, Italy, in 2003. Since 2000 she collaborates with the Intelligent Agents and Soft-Computing (IASC) Group at the Dept. of Electrical and Electronic Engineering (DIEE), University of Cagliari. Her educational background is mainly focused on intelligent agents, in particular on their proactive and adaptive behavior.

The above research topics are mainly experimented in information retrieval, text categorization, and bioinformatics.

The composition of ad-hoc agent-oriented design processes

Massimo Cossentino, ICAR-CNR, Sede di Palermo; Valeria Seidita, DINFO-Università degli Studi di Palermo

In the last years, the agent-oriented approach has been recognized as very suitable for the development of complex software systems since it fully exploits the well known techniques of Decomposition, Abstraction and Organization  for managing complexity.

The development of complex software systems by using the agent-oriented approach requires suitable agent-oriented modelling techniques and methodologies which provide explicit support for the key abstractions of the agent paradigm.

Several methodologies supporting analysis, design and implementation of Multi-Agent Systems (MASs) have been to date proposed in the context of Agent Oriented Software Engineering (AOSE) .

Although such methodologies have different advantages when applied to specific problems, it’s a matter of fact that an unique methodology cannot be general enough to be useful to everyone without some level of customization. In fact, agent designers, for solving specific problems in a specific application context, often prefer to define a new methodology specifically tailored for their needs instead of reusing an existing one.

Thus, an approach that combines the designer’s need of defining his own methodology with the advantages and the experiences coming from existing and well documented methodologies is highly required.

Biography: http://www.pa.icar.cnr.it/cossentino/

                    http://www.dinfo.unipa.it/index.php?page=seidita