Italy

Department of Electronics and Information, Politecnico di Milano

Membership Details (University/College)

Contact: Giuseppina Gini
Member node: 116
piazza L. da Vinci 32, 20133 Milano
Italy

>>> Include yourself at this organisation
>>> Update your name or personal details listed below
>>> Submit papers, projects etc to the archive
>>> Update Member Description or Contact details

Description

AGENTS AND PHYSICAL AGENTS

The aim of our research activity is the development of models, methodologies and tools for building adaptive physical agents and co-operative agents.

We have defined reinforcement learning models for both crisp and fuzzy robot controllers, and developed tools for implementing agents based on such models [1-5]. We are also working on experiments and formal definition of the aspects of an agency, i.e., a set of interacting agents. [6] Finally, we participate to the Robocop effort for soccer playing robots since the beginning.

Multi-robot systems are a perfect arena for multi-agent paradigm development.
The number of currently used robot is increasing and their level of development makes it possible to use them in new fields: recovering people caught in a risk area, surveying restricted areas, are examples where teams of professionals and robots can work together. In both these situations it is they are easier for groups of robots than for a single robot. The combination of the tele-operation and the techniques able to assure a high autonomy of the remote robot, named "tele-robotics" and often developed in supervisory control, are another agent related technology. The technologies of wearable computing, and ubiquitous computing, are also playing a role in our development. Any user can remotely interact with the robot as well as with other users.
Means of massive dissemination, like the Internet network, guarantees the ideal base for spreading these new technologies on a total scale. In our solutions client-server architectures allow the cooperation of different robots remotely controlled from different places. Each robot is a complex agent, equipped with sensors and controllers.
Another key idea is the direction of miniaturization that will give the possibility to the new agents to be on a chip and to become an extension of the human body. The chip could also integrate biological components, with cellular tissues to get better integration with the user. To this end we are developing artificial arts [7].

We have been and are working on those topics in Italian projects financed from from MURST and devoted to agent approaches to robotics and to teaching .

Another activity is the development of hybrid AI architectures for data analysis and data mining, where different components (agents) cooperate. See our work on mixtures of experts for predictive systems [8-10], developed in EU or NATO founded projects.

Article

Koening, C. and Gini, G. and Craciun, M. and Benfenati, E. (2004) Multi-class classifier from a combination of local experts: toward distributed computation for real-problem classifiers. International Journal of Pattern Recognition and Artificial Intelligence.

Bonarini, A. and Invernizzi, G. and Labella, T.H. and Matteucci, M. (2003) An architecture to coordinate fuzzy behaviors to control an autonomous robot. Fuzzy sets and systems, 134 (1). pp. 101-115.

Asada, M. and Obst, O. and Polani, D. and Browning, B. and Bonarini, A. and Fujta, M. and Christaller, T. and Takahashi, T. and Tadokoro, S. and Sklar, E. and Kaminka, G.A. and Eguchi, A. and Johnson, J. and Riley, P. and Hanek, R. and Schmitt, T. and Buck, S. and Beetz, M. (2003) An Overview of RoboCup 2002 Fukuoka/Busan. AI Magazine, Fall. pp. 21-40.

Mazzatorta, P. and Benfenati, E. and Neagu, C.-D. and Gini, G. (2003) Tuning Neural and Fuzzy-Neural Networks for Toxicity Modeling. Journal of Chemical Information and Computer Sciences, 43. pp. 513-518.

Bonarini, A. and Bonacina, C. and Matteucci, M. (2001) An approach to the design of reinforcement functions in real world, agent-based applications. IEEE Transactions on Systems, Man, and Cybernetics - Part B, 31 (3). pp. 288-301.

Bonarini, A. (2001) Evolutionary learning, reinforcement learning, and fuzzy rules for knowledge acquisition in agent-based systems. Proceedings of the IEEE, 89 (9). pp. 1334-1346.

Bonarini, A. and Trianni, V. (2001) Learning Fuzzy Classifier Systems for Multi-Agent Coordination. Information Sciences, 136 (1-4). pp. 215-239.

Amigoni, F. and Somalvico, M. and Zanisi, D. (1999) A Theoretical Framework for the Conception of Agency. International Journal of Intelligent Systems, 14 (5).

Book Section

Folgheraiter, M. and Gini, G. (2003) A Bio-inspired control system and a VRML simulator for an autonomous humanoid arm. In: Proc. IEEE Humanoids 2003. , Karlsruhe, Germany.

Benfenati, E. and Mazzatorta, P. and Neagu, C.-D. and Gini, G. (2002) Combining classifiers of pesticides toxicity through a neuro-fuzzy approach. In: MCS 2002. Lecture Notes in Computer Science (2364). Springer-Verlag, pp. 293-303.

Software

[Software] Picco, G.P. (2000) Mucode. [Other]

This list was generated on Mon Sep 10 16:39:27 BST 2007.

   

AgentLink is the European Commission's IST-funded Coordination Action for Agent-Based Computing
and is coordinated by the
University of Liverpool and University of Southampton
If you encounter any problems with these pages please contact web@agentlink.org.