D.2.3 NIAD&R: Ongoing projects |
This project is funded by the European Union IST Programme and supports an international network of excellence for Agent Based Computing. It follows AgentLink I and II existing for the last 6 years.
Project title: AgentLink III
Effort at LIACC: 1 man year
Funding entity: EU (IST)
Total award period covered: from September/2003
Coordinator at LIACC: Eugénio Oliveira
Prime contractor: University of Southampton
Other partners include, Queen Mary Westfield College/University of London, Technical University of Munich, University of York, OFAI (Vienna), University of Paris VI, University of Neuchatel, IIIA/Barcelona, Imag/Grenoble and many others.
Partners are grouped in different Special Interest Groups and NIAD&R - LIACC belongs to two different SIGs: Agent-Mediated Electronic Commerce (AMEC) and Adaptive and Learning Agents (ALAD). Our activity in AgentLink is directly related with our research on: Agents Negotiation protocol models
Our main objectives in this project are:
Project title: PORTUS (POSI/SRI/41315/2001)
Effort at LIACC: 71 man months
Duration period: 36 months
Funding entity: FCT, POSI
Coordinator at LIACC: Eugénio Oliveira
LIACC team: Luís Paulo Reis, Eugénio Oliveira
Partners: ISR-Porto (leader) and LIACC
Motivation: This project aims at developing a new approach to cooperative robotics in general and to building RoboSoccer teams in particular. The project main goal is to design a software common framework suitable for the implementation of an agent-based software that can control robots to be used for several cooperative robotic tasks and in particular, can control robots that are able to play in any Robotic Soccer leagues (simulation, small-size, middle-size and legged). The same agent-based software will control each one of the robots of Portus team, regarding its action and perception capabilities and the cooperative task to perform. The common framework will include a knowledge representation structure capable of representing structured knowledge supplied by a domain expert to a team of cooperative robots. This structure is instantiated by special supervisor agents developed for each RoboCup league or cooperative task to perform.
Each agent in Portus includes a dynamic world state model that will be updated though visual perception, robot communication and action prediction. In Portus project, both inter-agent cooperation and team coordination policies have to be investigated. This will include cooperation methods like dynamic positioning and role exchange, strategic positioning and intelligent communication.
In order to control different robots, the common framework software agent needs specific parts to deal with special constraints related with the agent's perception and action capabilities.
The very idea of Portus project is then to design an agent-based common framework applicable in partially cooperative, partially adversarial domains, to different cooperative robotic tasks, including all major RoboCup soccer leagues.
Main objectives of the project:
The main objective of Portus is to develop a common agent-based framework for controlling cooperative teams of mobile robots. According to the general architecture we are envisaging for the proposed framework, this objective may be subdivided into the following sub-objectives:
Portus main contribution will be a configurable common framework flexible enough to deal with perception, decision-making and action for different (both real and virtual) cooperative teams of robots.
Project title: LEMAS (POSI/ROBO/43926/2002)
Effort at LIACC: 18 man months
Duration period: 12 months
Funding entity: FCT, POSI
Coordinator at LIACC: Eugénio Oliveira
LIACC team: Luís Paulo Reis, Eugénio Oliveira, Ricardo Afonso, Ana
paula Rocha, Tiago Carvalho
Partners: Faculty of Engineering University of Porto, LIACC, ISR-Porto
Motivation: With LEMAS project we aim at developing a fully-functional "soccer" playing team of legged robots (AIBO "dogs" from Sony). Research will be focused on three main aspects: perception, autonomous decision-making and team coordination.
The first aspect implies developing a robust vision system including capabilities for colour image segmentation and object recognition. This sub-system will generate high-level descriptions of the image contents to be used on self-localization, action execution and cooperative playing.
It is also our intention to apply different kinds of reinforcement learning algorithms having in mind improving low-level skills of the legged robots.
Project title: FC Portugal (POSI/ROBO/43910/2002)
Effort at LIACC: 18 man months
Duration period: 12 months
Funding entity: FCT, POSI
Coordinator at LIACC: Luís Paulo Reis
LIACC team: Luís Paulo Reis, Eugénio Oliveira, Ricardo Afonso, Ana paula
Rocha, Tiago Carvalho
Partners: Faculty of Engineering University of Porto, LIACC, ISR-Porto
Motivation: FC Portugal - New Coordination Methodologies project intends to continue the research performed during the development of the successful FC Portugal RoboCup simulation league team. The main goal of this project research is the development of a formal model for the concept of team strategy for a competition with an opponent team having opposite goals, general enough to be instantiated to various dynamic competitive domains. The formal model will enable the design of an agent architecture suitable for RoboCup simulation league agents and a world state model capable of storing the information needed for an intelligent agent, to play soccer. The project research focus concerns on developing general decision-making and cooperation models for soccer playing. Cooperation mechanisms will include developments of the previously proposed Situation Based Strategic Positioning and Dynamic Positioning and Role Exchange Mechanisms. Communication languages and protocols, to convey the most relevant information at the right times to players will also be developed. Online optimization and reinforcement learning will be used in order to develop a complete set of efficient low-level skills for soccer playing agents. Coaching is an important research topic in RoboCup. We have proposed Coach Unilang - a general language to coach a (robo)soccer team. In this project, this language will be extended and fully implemented. Also, an agent architecture suitable for a simulation league coach and a complete soccer coaching ontology will be developed. FC Portugal will also be concerned with the development of agent evaluation tools like our offline client methodology; WstateMetrics that evaluates the accuracy of world states and Visual debugger used analyze the reasoning of agents. Evaluation by domain experts using graphical tools is another methodology that will be used. Finally, FC Portugal will participate in RoboCup 2004 that will be held in Lisbon - Portugal. This project intends to continue our successful participation in RoboCup and also to contribute to the development of other projects leading to the development of new teams that may participate in RoboCup 2004.
objectives: FC Portugal project main objective is the continuation of the development of FC Portugal simulation league team so that it successfully participates in RoboCup 2004 - Lisbon. The main objective is subdivided in several research objectives that enhance team competitiveness, namely:
Project title: Specification and monitoring of electronic contracts in
Virtual
Enterprises (PIDEEC/04-03/2003)
Effort at LIACC: 18 man months
Duration period: 12 months
Funding entity: DEEC-FEUP
Coordinator at LIACC: Ana Paula Rocha
LIACC team: Ana Paula Rocha, Daniel Moura, Henrique Lopes Cardoso
Partners: DEEC-Faculty of Engineering University of Porto
Motivation: In electronic markets, which are for definition open environments, there is the problem of establishing confidence between potential business partners (agents) and that are, generally, not known in advance. The traditional way to attribute security to a commercial transaction is through the establishment of a contract. The contract enumerates the dependencies between the partners that participate in the commercial activity. This project focus essentially in the phase of operation of the VE, and consists in the specification and monitoring of electronic contracts. The contracts are derived from templates that contain fixed and non-fixed parameters. These parameters that are not fixed will be the object of the negotiation process that occurs in the phase of formation of the VE (subject already explored in previous projects). Thus, the result of the VE formation phase will be the instantiation of an electronic contract to be used in the monitoring process that occurs during the VE operation phase. This contract is signed by all the VE partners, and possibly supervised for an independent agent. This project includes also a study about the language and the ontology of the contract.
D.2.3 NIAD&R: Ongoing projects |