Highlights of Practical Applications of Heterogeneous Multi-Agent Systems. The PAAMS Collection

The practical approach and multi-agent platform development for adaptive scheduling systems for real time resource management are considered. The approach is based on concept of demand and resource networks (DRN) where agents of demands and resources operate on virtual market and continuously trying to improve their individual values of satisfaction functions that reflects given multi-criteria objectives. To achieve the best possible results agents use the virtual money account that regulates their behavior and can increase by getting bonuses or decrease by penalties depending of their individual cost functions. The key rule of designed virtual market is that any agent that is searching for new better position in schedule must compensate losses for those conflicting agents who are able and agree to change their allocations to other resources after the initial agent request, with required amount of compensation determined in the process of re-allocations. This approach allows to balance many criteria for getting consensus between agents and adaptation of the schedules “on the fly” by events without any stop and restart of the system. The developed platform includes key classes of DRN agents and protocols of their negotiations and other components that help to develop the solution manage data and visualize results of scheduling. The platform provides rapid prototyping of multi-agent systems for real time resource management and helps to reduce man-efforts and time of development. The platform was applied for developing of multi-agent scheduling systems for managing resources in aircraft jet production, load balancing in computer grid networks and energy production in power-, gas- and heating networks.

Environmental issues, specifically pollution are considered as major concerns in many cities in the world. They have a direct influence on our health and quality of life. The use of computers models can help to forecast the impact of human activities on ecosystem equilibrium. We are interested in the use of MAS (Multi-Agent System) for modelling and simulating the environmental issues related to pollution. In this paper, we present a review of recent studies using a MAS approach for designing environmental pollution simulation models. Interactions between the three components of the environmental problem (Social, Economic and Ecological) are presented. On the light of these interactions, studies published from 2009 to 2013 are reviewed. Models are presented in terms of: model’s purpose, studied variables, used data, representation of space and time, decision-making mechanism and implementation.

The SESAR (Single European Sky ATM Research) program is an ambitious research and development initiative to design the future European air traffic management (ATM) system. The study of the behavior of ATM systems using agent-based modeling and simulation tools can help the development of new methods to improve their performance. This paper presents an overview of existing agent-based approaches in air transportation (paying special attention to the challenges that exist for the design of future ATM systems) and, subsequently, describes a new agent-based approach that we proposed in the CASSIOPEIA project, which was developed according to the goals of the SESAR program. In our approach, we use agent models for different ATM stakeholders, and, in contrast to previous work, our solution models new collaborative decision processes for flow traffic management, it uses an intermediate level of abstraction (useful for simulations at larger scales), and was designed to be a practical tool (open and reusable) for the development of different ATM studies. It was successfully applied in three studies related to the design of future ATM systems in Europe.

In recent years, intensive research and development have been done in the area of construction project risk management. Indeed, an efficient risk management is mandatory to project success. However, implementing such a management is complex because of the diversity and the dynamic nature of the risk. Moreover, each of the project stakeholders has his/her own risks, his/her own vision and his/her own action on the project and on risks. In this paper, we propose an agent-based model called SMACC to assess the impact of risks on the project. This model allows to test different risk mitigation strategies to measure their impact on the project. An application on a real project is also proposed to demonstrate the operability and the value of the proposed approach.

This article describes an agent which detects and handle potentially abnormal situations from the monitoring of applications usage on a tablet computer. The main purpose of this agent is to improve dependent people’s safety by signaling potentially risky situations to caregivers. Indeed, such signaling can improve response time, thus reducing the consequences of such situations. The detection of abnormal situations is based on the construction of a user profile from the monitoring of used applications. When a user is inactive during a certain period of time, the recent activity is compared to the learned user’s profile to decide if this is normal or not. Once an abnormal situation has been identified, the system will try to confirm that the situation is actually abnormal by prompting the user for input. In order to be as less intrusive as possible, the input request is an application suggestion. The suggested application will be the one that is usually the most used during the time period corresponding to the inactivity. When a situation is confirmed as abnormal, the tablet agent will send an intervention request to the user’s caregivers. A simple coordination mechanism aimed at reducing redundant interventions and improving caregivers response time is proposed. The main contribution of this work is to propose a mechanism which monitors elderly people’s applications usage on a tablet computer and is therefore able to complement existing monitoring devices in the detection of abnormal situations.

The feasibility of using a motion sensor to replace a conventional electrolarynx (EL) user interface was explored. A mobile phone motion sensor with multi-agent platform was used to investigate on/off and pitch frequency control capability. A very small battery operated ARM-based control unit was also developed to evaluate the motion sensor based user-interface. The control unit was placed on the wrist and the vibration device was placed against the throat using support bandage. Two different conversion methods were used for the forearm tilt angle to pitch frequency conversion: linear mapping method and F0 template-based method. A perceptual evaluation was performed with two well-trained normal speakers and ten subjects. The results of the evaluation study showed that both methods were able to produce better speech quality in terms of naturalness.

In this paper, we present an agent that is able to reason about the user’s emotions and to perform or suggest coping strategies to deal with them in order to improve the user’s well-being. Concretely, this agent uses the PLEAID reasoning engine that we extended to implement Dastani and Lorini’s BDI logic for graded emotions and coping strategies. We explain the difficulties behind such an implementation, that proves the computational tractability of the underlying logic. We then illustrate the possibilities offered by this agent on a short scenario involving an artificial companion interacting with a human user.

On government level, Denmark has published both strategies and technical guidelines to strengthen implementation and use of telehealth in the Danish healthcare sector in the future. Consequently telehealth solutions will become an integrated part of the daily life of the patients equipped with these solutions. This paper proposes an architecture for a multi-agent system to be implemented together with the telehealth solution in a patient’s home. The purpose of the multi-agent system is to incorporate more intelligence into the gathering of healthcare related data and thereby learn about the behavior and level of physical activity of the patient, and other interesting context information.

This paper proposes a cognitive architecture of an intelligent agent that can have a dialogue with a human agent on health-related topics. This architecture consists of four main components, namely, the

Belief Base

, the

Dialogue Manager

, the

Task Manager

and the

Plan Generator

. Each component has sub-components that perform a set of tasks for the purpose to enable the agent to be enrolled in a dialogue. In this paper the particular sub-component of the Dialogue Manager, the

Dialogue Strategy

has been discussed in detail. A notion of

scheme

is introduced, which functions as a template with variables that are instantiated each time a state is entered. The agent’s dialogue strategy is implemented as a combination of the schemes and the state transitions that the agent makes in response to the human’s request. We used a combination of finite-state and agent-based dialogue strategies for dialogue management. This combined dialogue strategy enables a multi-topic dialogue between a human and an agent.

Currently the Ambient Assisted Living and the Ambient Intelligence areas are very prolific. There is a demand of security and comfort that should be ensured at people’s homes. The AAL4ALL (ambient assisted living for all) project aims to develop a unified ecosystem and a certification process, allowing the development of fully compatible devices and services. The UserAccess emerges from the AAL4ALL project, being a demonstration of its validity. The UserAccess architecture, implementation, interfaces and test scenario are presented, along with the sensor platform specially developed for the AAL4ALL project.

For seniors who require continuous care and do not have the resources to have an assistant continuously, have a low cost system that monitors their environment allows them to have independence, while moving in a secure environment. In addition, accessing to basic services through a platform accessible to all people, including TV, facilitates their integration into the online society.

The paper presents a concept and application of a multi-agent approach to modeling and optimization the Multi-Echelon Capacitated Vehicle Routing Problem. Two environments (mathematical programming (MP) and constraint logic programming (CLP)) and two types of agents were integrated. The strengths of MP and CLP, in which constraints are treated in a different way and different methods are implemented, were combined to use the strengths of both. The proposed approach is particularly important for the decision models with an objective function and many discrete decision variables added up in multiple constraints. The Two-Echelon Capacitated Vehicle Routing Problem (2E-CVRP) is an extension of the classical Capacitated Vehicle Routing Problem (CVRP) where the delivery depot-customers pass through intermediate depots (called satellites). Multi-echelon distribution systems are quite common in supply-chain and logistic systems. The presented multi-agent approach will be compared with classical mathematical programming on the same data sets.

Today in the automotive industry, reconfigurability, flexibility and high availability are as important as the level of automation, cost effectiveness, and maximum throughput. In this paper two main delivery strategies (JIT – Just in Time, JIS – Just in Sequence) in the automotive industry are presented and two approaches to modeling and to simulate that strategies are described (DES - Discrete-Event and ABS - Agent-Based Simulation). Authors discuss that DES approach to JIS strategy is insufficient and propose to mixing DES and ABS in simulation of delivery processes in the automotive. The case study of proposed solutonion is described.

The aim of the paper is to present the agent based approach for modeling disturbances in supply chain. An introduction to the issue of Agent-Based Modeling is provided. The paper describes in detail the modeled area of a supply chain, taking into account different modeling techniques. Disturbances (selected and highlighted by the authors) are identified and modeling methods are discussed. Two selected disturbances are modeled using of agent-based approach. The research highlights of the performed works are as follows: identifying disturbances in a supply chain, which can be modeled with use of the ABS approach and demonstrating how to model the chosen disturbances by using the above method of modeling.

European economies have been deeply affected by different crises. The impact of the economic crisis on enterprises is now recognized by everybody. Enterprises need to reorganize in order to be better adapted to this situation. GRAI Methodology is one of the three main methodologies for enterprise modeling. GRAIMOD is a software tool being developed for supporting this methodology and facilitating enterprise improvement. The concepts elaborated for this tool combine reasoning like Case Based Reasoning (CBR), Decomposition or transformation reasoning and multi-agent systems like training agent. This paper introduces these concepts and presents how to complete them with simulation concepts for improving enterprise performance. An example will be used for illustrating the concepts presented through a detailed case study.

Multimodal dialog systems can be defined as computer systems that process two or more user input modes and combine them with multimedia system output. This paper is focused on the multimodal input, providing a proposal to process and fusion the multiple input modalities in the dialog manager of the system, so that a single combined input is used to select the next system action. We describe an application of our technique to build multimodal systems that process user’s spoken utterances, tactile and keyboard inputs, and information related to the context of the interaction. This information is divided in our proposal into external and internal context, user’s internal, represented in our contribution by the detection of their intention during the dialog and their emotional state.

Several crowd simulators simulate only homogenous pedestrians. In the reality, there are different personalities. This personality variation affects the pedestrian behavior and the fate of the crowd. In this paper, we extend the HuNAC (Human Nature of Autonomous Crowd) model by providing each pedestrian agent with a personality in order to examine how the emergent behavior of the crowd is affected. We use the MBTI personality theory as a basis for agent psychology. The aim of this work is to improve HuNAC model and consequently to improve results. In this context, we have a new version of our model which we called PHuNAC model (Personalities’ Human Nature of Autonomous Crowd). Our PHuNAC model is a multi-agent simulation of pedestrian crowd model .The conducted experiments show that PHuNAC model is able to produce more realistic pedestrian behaviors than the HuNAC model.

In this paper, we present a complete overview of Ambient Intelligence (AmI) focused in its applications, considering the involved domain and technologies. The applications include AmI at home, care of elderly and people with disabilities, healthcare, education, business, public services, leisure and entertainment. The aim of this survey of AmI’s applications is to show its socials and ethical implications and specially privacy issues. Intelligent Environments (IE) collect and process a massive amount of person-related and sensitive data. These data must ensure privacy of the users. An important concern in AmI´s applications is privacy. Addressing design by privacy, an important challenge to consider is the development of an architecture that includes the different privacy policies and how can we fusion them in a specific application domain. Ensuring privacy in Intelligent Environments is a difficult problem to solve, as there are different perceptions of privacy and its role in computing for each user. In the so called ‘design by privacy’ we have to identify the relevant design issues that should be addressed for its developing. Here we present an approach to the dimensions to consider, in order to provide privacy in the design of Ambient Intelligence’s applications.

Information fusion studies theories and methods to effectively combine data from multiple sensors and related information to achieve more specific inferences that could be achieved by using a single, independent sensor. Information fused from sensors and data mining analysis has recently attracted the attention of the research community for real-world applications. In this sense, the deployment of an Intelligent Offshore Oil Industry Environment will help to figure out a risky scenario based on the events occurred in the past related to anomalies and the profile of the current employee (role, location, etc.). In this paper we propose an information fusion model for an intelligent oil environment in which employees are alerted about possible risk situations while their are moving around their working place. The layered architecture, implements a reasoning engine capable of intelligently filtering the context profile of the employee (role, location) for the feature selection of an inter-transaction mining process. Depending on the employee contextual information he will receive intelligent alerts based on the prediction model that use his role and his current location. This model provides the big picture about risk analysis for that employee at that place in that moment.

Distributed energy resources (DER) offer an economically attractive alternative to traditional centralized generation. However, the unpredictable and scattered nature of DER prevents them from replacing traditional centralized generator capacity. Aggregating DER under a virtual power plant (VPP) addresses this issue by exposing the combined capabilities of connected DER as a single controllable entity towards the utility. In this paper we propose an architecture that supports multi-level aggregation of DER under VPPs. In this architecture, DER submit energy profiles to a VPP. The VPP may then control the DER within the boundaries defined in the energy profiles. The proposed architecture is hosted in an agent-based framework, Controleum, and is to be demonstrated at a primary school in Denmark, first quarter 2014.

Smart Grids are in focus of several international R&D past and present efforts since at least a decade. Smart Grids is a well-known metaphor for future power grids. However, the meaning, or semantics of the concept has, naturally, changed due to increased understanding of the inherent complexities of the subject matter. The driving forces behind the efforts on Smart Grids include:

Demands of integrating new energy sources such as Distributed Energy Resources (DER) and Renewable Energy Resources (RES) in a massive way into generation, transmission and distribution of future energy systems.

Establishment of a de-regulated customer oriented energy markets, including new types of energy based service markets.

Design and implementation of resilient and trustworthy services coordinating and monitoring use-case dependent sets of stakeholders during operations.

The transition from today’s mostly hierarchical power grids towards tomorrow’s Smart Grids poses several challenges to be properly addressed and harnessed. We argue that proper use of agent technologies is a key technology towards this end. Furthermore, we argue that design and implementation of Smart Grids have to be supported by Configurable Experiment Platforms to carter for the under specifications of such systems. Resilience of systems has several aspects. We focus on resilience related to different kinds of cyber attacks and self-healing.

The electricity networks in many countries are facing a number of challenges due to growth in peak demand, integration of renewable energy sources, increasing security risks and environmental concerns. Smart Grid, as an automated and widely distributed energy network, offers viable solutions to those challenges. Software agents running on customer premises or embedded in appliances and equipment can be used to plan future energy consumption and to shift loads according to pre-defined constraints. However, testing such distributed solutions prior to actual deployment in domestic households is a challenge. Simulations may not capture all the aspects of distributed, large-scale, complex environments, such as one can find in the Smart Grid. This paper presents a distributed Smart Grid simulation/emulation environment called Symphony that allows running real-world experiments within distributed environment with the participation of multiple actors. Symphony is being developed in the context of a European Institute for Innovation and Technology project.

This work proposes the use of a combination of gossip and consensus algorithms to allow a power grid to be self-organized by its components, adapting to the changes in the demand and compensating the possible failures that may occur. The Balearic Power Grid has been used to check the validity of the proposal.

The energy sector in industrialized countries has been restructured in the last years, with the purpose of decreasing electricity prices through the increase in competition, and facilitating the integration of distributed energy resources. However, the restructuring process increased the complexity in market players’ interactions and generated emerging problems and new issues to be addressed. In order to provide players with competitive advantage in the market, decision support tools that facilitate the study and understanding of these markets become extremely useful. In this context arises MASCEM (Multi-Agent Simulator of Competitive Electricity Markets), a multi-agent based simulator that models real electricity markets. To reinforce MASCEM with the capability of recreating the electricity markets reality in the fullest possible extent, it is crucial to make it able to simulate as many market models and player types as possible. This paper presents a new negotiation model implemented in MASCEM based on the negotiation model used in day-ahead market (Elspot) of Nord Pool. This is a key module to study competitive electricity markets, as it presents well defined and distinct characteristics from the already implemented markets, and it is a reference electricity market in Europe (the one with the larger amount of traded power).

Energy systems worldwide are complex and challenging environments. Multi-agent based simulation platforms are increasing at a high rate, as they show to be a good option to study many issues related to these systems, as well as the involved players at act in this domain. In this scope the authors’ research group has developed a multi-agent system: MASCEM (Multi-Agent System for Competitive Electricity Markets), which performs realistic simulations of the electricity markets. MASCEM is integrated with ALBidS (Adaptive Learning Strategic Bidding System) that works as a decision support system for market players. The ALBidS system allows MASCEM market negotiating players to take the best possible advantages from each market context. However, it is still necessary to adequately optimize the players’ portfolio investment. For this purpose, this paper proposes a market portfolio optimization method, based on particle swarm optimization, which provides the best investment profile for a market player, considering different market opportunities (bilateral negotiation, market sessions, and operation in different markets) and the negotiation context such as the peak and off-peak periods of the day, the type of day (business day, weekend, holiday, etc.) and most important, the renewable based distributed generation forecast. The proposed approach is tested and validated using real electricity markets data from the Iberian operator – MIBEL.

In competitive energy markets (EMs), customers can freely choose their energy suppliers. The electricity trade can be done in organized markets or using forward bilateral contracts. Currently, there are several simulation tools based on multi-agent techniques that allow modeling, partially or globally, competitive EMs. The existing tools allow simulating negotiation prices and volumes through bilateral contracts, transactions in pool markets, etc. However, these tools have some limitations, mainly due to the complexity of the electric system. In this context, this article focuses on bilateral trading and presents the key features of software agents able to negotiate forward bilateral contracts. Special attention is devoted to demand response in bilateral contracting, notably utility functions and trading strategies for promoting demand response. The article also presents a case study on forward bilateral contracting with demand response: a retailer agent and an industrial customer agent negotiate a 24h-rate tariff.

In competitive energy markets, customers can freely choose their energy suppliers. The electricity trade can be done in organized markets or using bilateral contracts between customers and suppliers. In the latter case, market participants set the terms and conditions of agreements independent of the market operator. They often enter into bilateral contracts to hedge against pool price volatility. Furthermore, these contracts are very flexible since the negotiating parties can specify their own contract terms. This article focuses on bilateral trading and presents the key features of software agents able to negotiate forward bilateral contracts. Special attention is devoted to risk management in bilateral contracting, notably utility functions and trading strategies for dealing with risk. The article also presents a case study on forward bilateral contracting involving risk management: a retailer agent and an industrial customer agent negotiate a 24h-rate tariff.

A Distributed Denial of Service attack (DDoS) is designed to overload a target device and its networks with packets to damage its resources or services. This paper proposes an Artificial Neural Network (ANN) detection engine to flag known and unknown attacks from genuine traffic. Based on experiments and data analysis, specific patterns are selected to separate genuine from DDoS packets, thus allowing normal traffic to reach its destination. The mitigation process is triggered when the detection system identifies attacks based on the known characteristic features (patterns) that were fed to the ANN during the training process. Such characteristic patterns separate attacks from normal traffic. We have evaluated our solution against related work based on accuracy, sensitivity, specificity and precision.

Few proposals exist in the literature where security in networks and communications is studied from a pro-active perspective. One of them is MARS, a self-healing system intended to mitigate the malicious effects of common threats in MANETs. MARS makes use of special agent nodes to recover the loss of connectivity due to the operation of malicious nodes in the environment. Despite the general good performance of MARS, this paper shows some situations in which it does not work properly. This is caused by an inappropriate behavior of the optimization objective function considered in MARS. To overcome this limitation, a couple of alternative functions are designed and evaluated. The effectiveness of the new proposals is validated through extensive experiments. The new optimization functions lead to an increase in the resilience and tolerance of the network against security threats, improving network survivability.

There are many trials where electronic evidences are not digitally signed. This problem requires the validation of the veracity of these digital resources, thus causing delays in the verdict and increase the price of the process. The digital signature solves partially the problem because it provides the characteristics of authentication, integrity and non-repudiation; but it has weakness such as availability, confidentiality and changes control, besides the complexity on its usage. This paper presents the project DoyFE.es that is an agent-based platform deployed over a cloud system that provided cloud-based services to guarantee the veracity of the communications (email, web content and photographs).

Agreement technologies have settled the basis for creating systems that operate on the basis of agreements in societies of independent, autonomous computational entities (agents). However, nowadays more and more systems of such kind rely on a seamless interaction of software agents with humans. Humans work in partnership (directly or indirectly) or closely related with agents that are able to act autonomously and intelligently. Specifically, humans and agents have the ability to establish a series of relationships/collaborative interactions with each other, forming what might be called human-agent teams to meet their individual or collective goals within an organisation or social structure. Systems in which people and agents operate on a large scale offer an enormous potential but also require the consideration of additional issues. In this paper we analyse the open issues that may be addressed for researches in order to develop open human-agent systems. We present a real-world case study and an abstract architecture proposal for such systems.

Multi-agent technology allows the development of current AmI applications. Specifically, a multi-agent system allows the formation and management of applications where the main components can be humans and software agents interact and communicate with humans in order to help them in their daily activities. This kind of applications are what we call a

Human-Agent Society

, where agents provide services to humans or to other agents in an environment of whole integration. This paper presents a solution for the problem of human immersion presented in this kind of systems, providing the use of natural interfaces for the interaction among humans and software agents.

One of the most important factors to determine the success of an organization is the quality of decisions made. In order to improve the decisions taken and to strengthen the competitiveness of organizations, systems such as Group Decision Support Systems (GDSSs) have been strongly developed and studied in recent decades. The amount of GDSSs incorporating automatic negotiation mechanisms, such as argumentation, is increasing nowadays. The evaluation of these mechanisms and the understanding of their real benefits for the organizations is still a hard challenge. In this article, we propose a model that allows a GDSS to measure the participant’s satisfaction with the decision, considering aspects such as problem evaluation, personality, emotions and expectations. This model is intended to enable the understanding of the decision’s quality achieved with an argumentation system and to evaluate its capability to potentiate the decision’s quality. The proposed model validates all the assumptions found in the literature regarding the participant’s satisfaction.

The specialization exists in biological systems and in human organizations, as a methodology to improve processes and optimize their aims. This specialization in artificial intelligent systems such as multi-agent systems, can improve their aims, depending on the type of specialization and the goals which they need to achieve. The enterprise networks are a collaboration model between companies which we can apply over these intelligent systems, so that, these systems can achieve more complex aims. Therefore, in this collaboration type, is necessary to consider their specialization type, and how they could collaborate to achieve aims, that by themselves would not be possible.