Paulo Roberto Ferreira

Scheduling meetings through multi-agent negotiations

This work presents a set of protocols for scheduling a meeting among agents that represent their respective users interests. Four protocols are discussed: a) the full information protocol when all agents are comfortable with sharing their preference profile and free times; b) the approval protocol when only the preference profile can be shared; c) the voting protocol when only free time can be shared; and d) the suggestion protocol if neither preference nor free time can be shared. We use non-standard metric to evaluate the protocols which aims at maximizing the average preference, but also seeks to reduce the differences in preferences among the agents. The full information and approval protocols are optimal, that is, they achieve the best solution. Results show that the voting protocol achieves the best solution 88% of the time. Simulation results for the suggestion protocol with different numbers of agents, different numbers of solutions, and different strategies are presented. The suggestion protocol is shown to be coalition-free.

RoboCup Rescue as multiagent task allocation among teams: experiments with task interdependencies

This paper addresses distributed task allocation among teams of agents in a RoboCup Rescue scenario. We are primarily concerned with testing different mechanisms that formalize issues underlying implicit coordination among teams of agents. These mechanisms are developed, implemented, and evaluated using two algorithms: Swarm-GAP and LA-DCOP. The latter bases task allocation on a comparison between an agent’s capability to perform a task and the capability demanded by this task. Swarm-GAP is a probabilistic approach in which an agent selects a task using a model inspired by task allocation among social insects. Both algorithms were also compared to another one that allocates tasks in a greedy way. Departing from previous works that tackle task allocation in the rescue scenario only among fire brigades, here we consider the various actors in the RoboCup Rescue, a step forward in the direction of realizing the concept of extreme teams. Tasks are allocated to teams of agents without explicit negotiation and using only local information. Our results show that the performance of Swarm-GAP and LA-DCOP are similar and that they outperform a greedy strategy. Also, it is possible to see that using more sophisticated mechanisms for task selection does pay off in terms of score.

A Swarm-Based Approach for Selection of Signal Plans in Urban Scenarios

This paper presents a swarm approach to the problem of synchronisation of traffic lights in order to reduce traffic jams in urban scenarios. Other approaches for reducing jams have been proposed. A classical one is to coordinate or synchronise traffic lights so that vehicles can traverse an arterial in one direction, with a specific speed, without stopping. Coordination here means that if appropriate signal plans are selected to run at the adjacent traffic lights, a “green wave” is built so that drivers do not have to stop at junctions. This approach works fine in traffic networks with defined traffic flow patterns like for instance morning flow towards downtown and its similar afternoon rush hour. However, in cities where these patterns are not clear, that approach may not be effective. This is clearly the case in big cities where the business centres are no longer located exclusively downtown.

Using Swarm-GAP for Distributed Task Allocation in Complex Scenarios

This paper addresses distributed task allocation in complex scenarios modeled using the distributed constraint optimization problem (DCOP) formalism. It is well known that DCOP, when used to model complex scenarios, generates problems with exponentially growing number of parameters. However, those scenarios are becoming ubiquitous in real-world applications. Therefore, approximate solutions are necessary. We propose and evaluate an algorithm for distributed task allocation. This algorithm, called Swarm-GAP, is based on theoretical models of division of labor in social insect colonies. It uses a probabilistic decision model. Swarm-GAP is experimented both in a scenario from RoboCup Rescue and an abstract simulation environment. We show that Swarm-GAP achieves similar results as other recent proposed algorithm with a reduction in communication and computation. Thus, our approach is highly scalable regarding both the number of agents and tasks.

A Swarm Based Approach for Task Allocation in Dynamic Agents Organizations

One of the well-studied issues in multi-agent systems is the standard action-selection and sequencing problem where a goal task can be performed in different ways, by different agents.Tasks have constraints while agents have different characteristics such as capacity, access to resources, motivations, etc. This class of problems has been tackled under different approaches. Moreover, in open, dynamic environments, agents must be able to adapt to the changing organizational goals, available resources, their relationships to another agents, and so on. This problem is a key one in multi-agent systems and relates to models of learning and adaptation, such as those observed among social insects. The present paper tackles the process of generating, adapting, and changing multiagent organization dynamically at system runtime, using a swarm inspired approach. This approach is used here mainly for task allocation with low need of pre-planning and specification, and no need of explicit coordination. The results of our approach and another quantitative one are compared here and it is shown that in dynamic domains, the agents adapt to changes in the organization, just as social insects do.

Optimizing Preferences within Groups: A Case Study on Travel Recommendation

This work describes a multiagent recommender system where agents work on behalf of members of a group of customers, trying to reach the best recommendation for the whole group. The goal is to model the group recommendation as a distributed constraint optimization problem, taking customer preferences into account and searching for the best solution. Experimental results show that this approach can be sucessfully applied to propose recommendations to a group of users.

A swarm based approximated algorithm to the extended generalized assignment problem (E-GAP)

This paper addresses distributed task allocation in complex scenarios modeled using the distributed constraint optimization problem (DCOP) formalism. We propose and evaluate a novel algorithm for distributed task allocation based on theoretical models of division of labor in social insect colonies, called Swarm-GAP. Swarm-GAP was experimented in an abstract centralized simulation environment and in the RoboCup Rescue Simulator. We show that Swarm-GAP achieves similar results to other recent proposed algorithm with a dramatic reduction in communication and computation. Thus, our approach is highly scalable regarding both the number of agents and tasks.

A swarm based approach to adapt the structural dimension of agents' organizations

One of the well studied issues in multi-agent systems is the standard action-selection problem where a goal task can be performed in different ways, by different agents. Also the sequence of these actions can influence the goal achievement or its quality. This class of problems has been tackled under different approaches. At the high-level coordination one, the specification of the organizational issues is crucial. However, in dynamic environments, agents must be able to adapt to the changing organizational goals, available resources, their relationships to the presence of another agents, and so on. This problem is a key one in multi-agent systems and relates to models of learning and adaptation, such as those observed among social insects. The present paper tackles the process of generating, adapting, and changing multi-agent organization dynamically at system runtime, using a swarm inspired approach. This approach is used here mainly for task allocation with low need of pre-planning and specification, and no need of explicit coordination. The results of our approach and another quantitative one are compared here and it is shown that in dynamic domains, the agents adapt to changes in the organization, just as social insects do.

Action Selection and Sequencing in Multiagent Systems: An Approximate Algorithm Based on Swarm Intelligence

One of the well studied issues in multiagent systems is the action-selection and sequencing problem where a goal is decomposed in tasks that can be performed in different ways and/or by different agents. This problem has been tackled under different approaches. In particular, for open, dynamic environments agents must be able to adapt to the changing organizational goals, available resources, their relationships to another agents, and so on. This problem is a key one in multi-agent systems and relates to models of adaptation, such as those observed among social insects. This paper shows how mechanisms from Swarm Intelligence are used to solve the action-selection and sequencing problem in dynamically changing environments that can have large number of agents and tasks.

Applying a Distributed Swarm-Based Algorithm to Solve Instances of the RCPSP

This paper addresses distributed task scheduling problems generalized as a distributed version of the Resource-Constrained Project Scheduling Problem (RCPSP) [1]. We apply and evaluate a novel approach for the RCPSP that is distributed and based on theoretical models of division of labor in social insects.