Industry Track
Session A
Over the course of the 20th century, the electrical powersystems of industrialized economies have become one of themost complex systems created by mankind. A number ofongoing trends will drastically change the way this critical infrastructure is operated. Demand for electricity keepsgrowing while the controllability of generation capacity isdecreasing due to introduction of renewable energy sources.Further, there is an increase of distributed generators (DG),i.e. the generation capacity embedded in the (medium andlow voltage) distribution networks. Intelligent distributedcoordination will be essential to ensure the electricity infrastructure runs efficiently in the future. The PowerMatchertechnology, a multi-agent coordination system, has been developed to provide this kind of coordination. The heartof the system is an electronic market on which local control agents negotiate using strategies based on short-termmicro-economics. This concept has been demonstrated ina number of field tests of increasing scale. Currently, thefocus is moving from proof of concept field tests proving thetechnology towards demonstrations developing commercialapplications paving the way for large-scale application. Multi-Agent Coordination in the Electricity Grid, from Concept towards Market Introduction
The trend towards renewable, decentralized, and highly fluctuating energy suppliers (e.g. photovoltaic, wind power, CHP)introduces a tremendous burden on the stability of futurepower grids. By adding sophisticated ICT and intelligentdevices, various Smart Grid initiatives work on concepts forintelligent power meters, peak load reductions, efficient balancing mechanisms, etc. As in the Smart Grid scenario datais inherently distributed over different, often non-cooperativeparties, mechanisms for efficient coordination of the suppliers, consumers and intermediators is required in order toensure global functioning of the power grid. In this paper,a highly flexible market platform is introduced for coordinating self-interested energy agents representing power suppliers, customers and prosumers. These energy agents implement a generic bidding strategy that can be governedby local policies. These policies declaratively represent userpreferences or constraints of the devices controlled by theagent. Efficient coordination between the agents is realizedthrough a market mechanism that incentivizes the agents toreveal their policies truthfully to the market. By knowingthe agent’s policies, an efficient solution for the overall system can be determined. As proof of concept implementationthe market platform D’ACCORD is presented that supportsvarious market structures ranging from a single local energyexchange to a hierarchical energy market structure. An Agent-based Market Platform for Smart Grids
We propose a multiagent based interpolation system for traffic conditions that includes estimation and learning agents.These agents are allocated to all the road links. The Normalized Velocity (NV) is used in this system. Estimation agentsrenew the NV for each road link, and learning agents renewthe weight values for estimation. The weight values can becalculated by multivariate analysis. Estimation and learning agents alternately calculate the results to improve theinterpolation accuracy. The Coefficient of Determination(CD) and Mean Square Error (MSE) are used to evaluatethe interpolation accuracy. Vehicle Information and Communication System (VICS) data and Probe Car Data (PCD)are usually used for traffic information systems, but we haveconfirmed that the estimation accuracy without VICS data(only PCD) is higher than with VICS data. The standarddeviation of the estimated NV error can be improved to0.1312, and the standard deviation of the estimated velocityerror is 6.56 km/h in the mid velocity region. It was possible to improve the CD and MSE by repeated estimationand learning. Multiagent Based Interpolation System for Traffic Condition by Estimation/Learning
Wireless cognitive radio (CR) is a newly emerging paradigmthat attempts to opportunistically transmit in licensed frequencies, without affecting the pre-assigned users of thesebands. To enable this functionality, such a radio must predict its operational parameters, such as transmit power andspectrum. These tasks, collectively called spectrum management, is difficult to achieve in a dynamic distributed environment, in which CR users may only take local decisions, andreact to the environmental changes. In this paper, we introduce a multi-agent reinforcement learning approach basedspectrum management. Our approach uses value functionsto evaluate the desirability of choosing different transmission parameters, and enables efficient assignment of spectrums and transmit powers by maximizing long-term reward. We then investigate various real-world scenarios, andcompare the communication performance using different setsof learning parameters. We also apply Kanerva-based function approximation to improve our approach’s ability to handle large cognitive radio networks and evaluate its effect oncommunication performance. We conclude that our reinforcement learning based spectrum management can significantly reduce the interference to the licensed users, whilemaintaining a high probability of successful transmissions ina cognitive radio ad hoc network. Spectrum Management of Cognitive Radio Using Multi-agent Reinforcement Learning
An ever-growing infrastructure, including existing and newlybuilt power plants, as well as a rising environmental awareness in society call for inspection and maintenance systemsof high efficiency. A solution can be found in the development of mobile agents to provide assistive inspection toolswith improved autonomy. In collaboration with industry theMagneBike robot for power plant inspection has been developed. The robot has been tested in a specific real field environment showing critical issues but inspiring future guidelines. This paper proposes to turn the semi-autonomousMagneBike robot into a multi-agent inspection system withclear benefits in speed, robustness and flexibility of task execution. The inspection task is approached by a hybrid coverage method that combines the concepts of blanket and sweepcoverage. Three algorithms implementing hybrid coverageare presented and evaluated in simulations. MagneBike - Toward multi climbing robots for power plant inspection
Session B
Most of women’s deaths related to pregnancy occur in newlyindustrialized countries. In association with gynecologistsand obstetricians of the Ant\^{o}nio Pedro University Hospitalo(HUAP) in Brazil, we have identified deficiencies in the prenatal care of the Brazilian public healthcare system that canbe computer-supported. They are mainly related to protocols that must be followed in the primary healthcare institutions and the referral process that must take place when ahigh risk pregnancy is identified, besides other functionalities that can be automated by a software application. In thispaper, the Prenatal Care Unified System (SUAP) projectwill be introduced, which provides a Multi-agent System forsupporting and monitoring the prenatal care. This projectuses agent technology to manage healthcare records, to actas a clinical decision support system, and to handle the logistics of high risk pregnancy cases. We also describe the challenges encountered during the implementation of the SUAPand discuss the benefits that an agent-based solution provided to the development of our system. Supporting Prenatal Care in the Public Healthcare System in a Newly Industrialized Country
Commercial aviation transportation is on the rise and hasbecome a necessity in our increasingly global world. Thereis a societal demand for more options, more traffic, moreefficiency, while still maintaining safety in the airspace. Tomeet these demands the Next Generation Air Transportation System (NextGen) concept from NASA calls for technologies and systems offering increasing support from automated decision-aiding and optimization tools. Such systemsmust coordinate with the human operator to take advantage of the functions each can best perform: The automatedtools must be designed to support the optimal allocation oftasks (functions) between the system and the human operators using these systems. Preliminary function allocationmethods must be developed (and evaluated) that focus onthe NextGen Airportal challenges, given a flexible, changingConcept of Operations (ConOps).We have begun making steps toward this by leveragingwork in agents research (namely Adjustable Autonomy) inorder to allow function allocation to become more dynamicand adjust to the goals, demands, and constraints of thecurrent situation as it unfolds. In this paper we introduceDynamic Function Allocation Strategies (DFAS) that arenot static and singular, but rather are represented by allocation policies that vary over time and circumstances. TheNextGen aviation domain is a natural fit for agent basedsystems because of its inherently distributed nature and theneed for automated systems to coordinate on tasks mapswell to the adjustable autonomy problem. While current adjustable autonomy methods are applicable in this context,crucial extensions are needed to push the existing models tolarger numbers of human players, while maintaining criticaltiming. To this end, we have created an air traffic controlsystem that includes: (1) A simulation environment, (2) aDFAS algorithm for providing adjustable autonomy strategies and (3) the agents for executing the strategies and measuring system efficiency. We believe that our system is thefirst step towards showing the efficacy of agent supportedapproach to driving the dynamic roles across human operators and automated systems in the NextGen environment.We present some initial results from a pilot study using thissystem. Function Allocation for NextGen Airspace via Agents
Safety culture is broadly recognized as important for Air TrafficManagement and various studies have addressed itscharacterization and assessment. Nevertheless, relations betweensafety culture and formal and informal organizational structuresand processes are yet not well understood. We aim to improve theunderstanding of these relations by agent-based organizationalmodeling and thus provide a way for structured improvement ofsafety culture. This paper presents the key elements, results andvalidation of an agent-based organizational model for a particularAir Navigation Service Provider. Can We Predict Safety Culture?
Room clearing, in which building surveillance is conducted tosearch for criminals, continues to be a dangerous and difficultproblem in urban settings, for both the military as well as forpolice. In a typical setting, an unknown number of hostile forcesmay be located in a building, and they may be armed.Furthermore, there may be innocent civilians. The goal of thefriendly units is to enter the room and secure it, but without lossof life of friendly forces, hostile forces, and most especially ofinnocent civilians. It would be beneficial to allow robots to be apart of the friendly team, however it is very challenging to haverobots that do not either slow down or obstruct their humanteammate. This is especially difficult since nearly all robots in useby the military and police today are tele-operated. In this paper,we describe work we have developed in cooperation with thearmy, for the room clearing domain. We constructed an algorithmwhereby multiple agents, in the form of robots, can accomplish aroom clearing task. We augmented the agent algorithms tointroduce Adjustable Autonomy, allowing cooperation withhumans. We describe simulated results of the algorithm onbuilding maps, and furthermore we describe how we intend tonext conduct hardware tests, and eventual plans to field thesystem. This agent-based solution has great potential to increasethe acceptance and leverage of robotics in complex environments. Agent-based Coordination of Human-Multirobot Teams in Complex Environments
Issue tracking is an essential part of regulated software development where it is typically supported by software systemswhich are complex and not easily customizable. We proposea meta-software agent that senses what windows and widgets are in focus by the user and leverages this awarenessto provide support. The user is given ways of making andrecalling annotations appropriate for the context. By observing users in action the agent creates models which canthen be used to predict and suggest next steps. This paper describes an early prototype of this approach built asa proof of concept. Preliminary results and directions forfuture work are outlined. MAITH: a Meta-software Agent for Issue Tracking Help