Multi-agent Teamwork Research Articles

Dynamic role discovery and assignment in multi-agent task decomposition

Effective multi-agent teamwork can be facilitated by using personas to decompose goals into lower-level team subtasks through a shared understanding of multi-agent tasks. However, traditional methods for role discovery and assignment are not scalable and fail to adapt to dynamic changes in the environment. To solve this problem, we propose a new framework for learning dynamic role discovery and assignment. We first introduce an action encoder to construct a vector representation for each action based on its characteristics, and define and classify roles from a more comprehensive perspective based on both action differences and action contributions. To rationally assign roles to agents, we propose a representation-based role selection policy based on consideration of role differences and reward horizons, which enables agents to play roles dynamically by dynamically assigning agents with similar abilities to play the same role. Agents playing the same role share their learning of the role, and different roles correspond to different action spaces. We also introduce regularizers to increase the differences between roles and stabilize training by preventing agents from changing roles frequently. Role selection and role policy integrate action representations and role differences in a restricted action space, improving learning efficiency. We conducted experiments in the SMAC benchmark and showed that our method enables effective role discovery and assignment, outperforming the baseline on four of the six scenarios, with an average improvement in win rate of 20%, and is effective in hard and super hard maps. We also conduct ablation experiments to demonstrate the importance of each component in our approach.

Engineering Challenges Ahead for Robot Teamwork in Dynamic Environments

The increasing number of robots around us creates a demand for connecting these robots in order to achieve goal-driven teamwork in heterogeneous multi-robot systems. In this paper, we focus on robot teamwork specifically in dynamic environments. While the conceptual modeling of multi-agent teamwork was studied extensively during the last two decades and commercial multi-agent applications were built based on the theoretical foundations, the steadily increasing use of autonomous robots in many application domains gave the topic new significance and shifted the focus more toward engineering concerns for multi-robot systems. From a distributed systems perspective, we discuss general engineering challenges that apply to robot teamwork in dynamic application domains and review state-of-the-art solution approaches for these challenges. This leads us to open research questions that need to be tackled in future work.

Extreme teams: Toward a greater understanding of multiagency teamwork during major emergencies and disasters.

Major emergencies are extreme team decision making environments. They are complex, dynamic, high-stakes and fast paced events, wherein successful resolution is contingent upon effective teamwork. Not only do emergency teams coordinate at the intrateam level (e.g., police team), but they are increasingly required to operate at the interteam level (e.g., police, fire and ambulance teams). This is in response to the desire for networked and cost-effective practice and due to the evolving nature of modern threats, such as extreme weather events and terrorist attacks, which require a multi- rather than single-agency response. Yet the capacity for interoperability between emergency teams is under researched and poorly understood. Much of the teamwork research is based on student-samples or in artificial lab settings, reducing the salient contextual demands of emergencies (e.g., high-stakes, meaningful risk). Furthermore, the minimal research that has been conducted has tended to provide broad descriptive accounts of challenges faced during emergencies, but failed to develop and test solutions. This article identifies what is known about emergency teams and highlights why it is an important and timely area for research. It will focus on the challenges and solutions to three areas of team processing: cooperation, coordination, and communication. Future research must have a solutions-focused approach. This can be oriented around areas: training, sociotechnical networks, and policies/procedural guidelines. Greater collaboration between academics and practitioners can grow knowledge in this domain, ensuring that interventions to improve emergency teamwork are both contextually grounded and empirically validated. (PsycINFO Database Record

Multi-agency working and implications for care managers

Purpose To examine the association between multi-agency working and psychosocial characteristics of work, practitioner time-use and job satisfaction. Design/methodology/approach A comparison of practitioners working in multi-agency (health and social care) and single-agency (social care only) teams, using data from the 2008 evaluation of Individual Budgets pilots in England. Participants worked in care manager roles supporting adult social care service users, and comprised social workers and a smaller number of health professionals. Data was collected using a self-completed questionnaire. Findings Data were returned from 249 respondents (a 29 per cent response rate), with two-thirds working in single-agency teams. No significant differences were found between team type and job satisfaction. Respondents in multi-agency teams reported greater decision autonomy but poorer supervisory support than those in single-agency teams. The latter finding was robust to further exploration using regression to control for confounding factors. Research limitations/implications These data were not specifically collected for the study and response rates were relatively low due to organisational upheaval at the time of data collection, which may affect interpretation. Practical implications Government policy is dedicated to extending integrated forms of working, including multi-agency teamwork. This research suggests that such structures need careful planning for them to work effectively, with particular attention to supervision arrangements. Originality/value This research gives a systematic and objective exploration of the relationship between job characteristics, time-use and satisfaction of practitioners in single as compared to multi-agency teams.

An extended study on addressing defender teamwork while accounting for uncertainty in attacker defender games using iterative Dec-MDPs

Multi-agent teamwork and defender-attacker security games are two areas that are currently receiving significant attention within multi-agent systems research. Unfortunately, despite the need for effective teamwork among multiple defenders, little has been done to harness the teamwork research in s ecurity games. The problem that this paper seeks to solve is the coordination of decentralized defender agents in the presence of uncertainty while securing targets against an observing adversary. To address this problem, we offer the following novel contributions in this paper: (i) New model of security games with defender teams that coordinate under uncertainty; (ii) New algorithm based on column generation that utilizes Decentralized Markov Decision Processes (Dec-MDPs) to generate defender strategies that incorporate uncertainty; (iii) New techniques to handle global events (when one or more agents may leave the system) during defender execution; (iv) Heuristics that help scale up in the number of targets and agents to handle real-world scenarios; (v) Exploration of the robustness of randomized pure strategies. The paper opens the door to a potentially new area combining computational game theory and multi-agent teamwork.

Reorganizing Complex Network to Improve Large-Scale Multiagent Teamwork

Large-scale multiagent teamwork has been popular in various domains. Similar to human society infrastructure, agents only coordinate with some of the others, with a peer-to-peer complex network structure. Their organization has been proven as a key factor to influence their performance. To expedite team performance, we have analyzed that there are three key factors. First, complex network effects may be able to promote team performance. Second, coordination interactions coming from their sources are always trying to be routed to capable agents. Although they could be transferred across the network via different paths, their sources and sinks depend on the intrinsic nature of the team which is irrelevant to the network connections. In addition, the agents involved in the same plan often form a subteam and communicate with each other more frequently. Therefore, if the interactions between agents can be statistically recorded, we are able to set up an integrated network adjustment algorithm by combining the three key factors. Based on our abstracted teamwork simulations and the coordination statistics, we implemented the adaptive reorganization algorithm. The experimental results briefly support our design that the reorganized network is more capable of coordinating heterogeneous agents.

Teamwork with Limited Knowledge of Teammates

While great strides have been made in multiagent teamwork, existing approaches typically assume extensive information exists about teammates and how to coordinate actions. This paper addresses how robust teamwork can still be created even if limited or no information exists about a specific group of teammates, as in the ad hoc teamwork scenario. The main contribution of this paper is the first empirical evaluation of an agent cooperating with teammates not created by the authors, where the agent is not provided expert knowledge of its teammates. For this purpose, we develop a general-purpose teammate modeling method and test the resulting ad hoc team agent's ability to collaborate with more than 40 unknown teams of agents to accomplish a benchmark task. These agents were designed by people other than the authors without these designers planning for the ad hoc teamwork setting. A secondary contribution of the paper is a new transfer learning algorithm, TwoStageTransfer, that can improve results when the ad hoc team agent does have some limited observations of its current teammates.

An Architecture of Intelligent Multi-Agent System: a Case Study in Soccergame Simulation

Multi-agent collaboration or teamwork and learning are critical research challenges especially for a large number of multi-agent applications. In multi-agent system, agents must accurately perceive and act despite a quickly changing and noisy environment. They have to act not only at individual skills but also at collaborative and adversarial behaviours. Soccer team simulations are used to provide a suitable test bed for research in multi-agent collaboration systems. In this paper, we proposed agent architecture for soccer team simulation. It is an agent architecture within which a flexible team organization and decision-making are presented. The proposed agent architecture includes four main components. First, the world model that describes the environment, which provides a very important resource for making decisions. The world model information is collected through environment monitoring and world prediction for the unobserved information. The second component is team organization, which is achieved by means of “Team Agreement”. The third one is an inter-agent communication paradigm that enables team coordination and knowledge transfer between agents. The fourth is a decision-making process, which includes passing policy, shoots policy, and strategic position policy. In the decision-making process, we proposed and implemented heuristic passing algorithm, neural-network-based shooting algorithm, and heuristic strategic position algorithm. Experimental results of the proposed system prove its stability and applicability. Hence more, the experimental results conclude that coaching enhances the proposed collaboration multi-agent system.

Risk and professionalism

In uncertain times for psychiatrists and for the medical profession in general, defining roles and responsibilities is proving to be a complex task. Yet, as clinicians grapple with multidisciplinary and multiagency teamwork, one expectation of consultant psychiatrists has become clear – their

Developing Multiprofessional Teamwork for Integrated Children's Services

By Anning, Angela, Cottrell, David, Frost, Nick, Green, Josephine and Robinson, Mark Maidenhead : Open University Press , 2006 ISBN 9780335219780 , 147 pp, £17.99 (pb) Despite the strong current emphasis on the value of multi-agency working, there is still little research evidence on the impact of joined up services and multi-professional practice on outcomes for service users. It was within this context — and before the publication of Every Child Matters or the 2004 Children Act — that the research project which informs this book was begun. Funded by the Economic and Social Research Council and based at Leeds University, the Multi-Agency Teamwork for Children's Services (MATCh) project was set up to explore, over a 2-year period, the daily realities of delivering public and voluntary sector services by multi-agency teamwork. The project worked with five well-established teams, exemplifying the type of multi-professional teams operating across the statutory and voluntary sectors — a youth offending team, a young people's Child and Adolescent Mental Health team, a nursery for children with learning difficulties, a head injury team and a child development team. Using two theoretical frameworks —‘communities of practice’ and activity theory — the study examines how the teams developed and functioned, how they were organised and managed, and the experiences of the individuals working in them. The issues that are raised will be familiar to anyone who has worked with colleagues with different training, beliefs, value systems and language and who are likely to hold different constructs of family and childhood. Working in inter-agency teams challenges professional identities and status and leads to a new sense of ‘who I am’. The authors also point to the way in which the nature of professional knowledge is challenged —‘what I know’ and ‘what I do’ draws on a combination of formal and informal, and theoretical and experiential knowledge, and whilst much of this knowledge is tacit within the workplace, in multi-agency teams it has to be made more explicit and drawn from across the team. The point is also made that it remains important to respect and deploy distinctive specialisms within teams, as well as general understanding, if professionals are to gain job satisfaction. The main value of this book for managers and practitioners is its discussion of some of the key dilemmas faced by multi-professional teams. The researchers introduced a series of scenarios to the members of their teams, illustrating specific incidents that arose during the course of the project and the dilemmas they presented — for example about sharing and acknowledging specialist expertise, reshaping working practices, negotiating new ways of working, resolving difference of values, transferring skills, sharing information, etc. They draw on the response to these scenarios, and on other research in the field, to explore strategies for resolving these dilemmas, focusing in particular on strategies for making decisions and delivering services for children. A number of themes are examined, amongst them shared procedures, clear lines of accountability, clarity on employment conditions, effective leadership, agreed strategic objectives and shared aims, transparent communication structures and ongoing support for professional development. This is a timely book and one that will be of value to all those who work across agencies to plan, commission, deliver and review services for children and families. But we should not forget that the main purpose of multi-professional and across-agency working is to deliver services that are more effective in meeting the needs of children and improving their well-being. Multi-professional working is a means to an end, and an important means, but not an end in itself.

Observation-Based Proactive Communication in Multi-Agent Teamwork

Multi-agent teamwork is governed by the same principles that underlie human cooperation. This paper describes how to give agents the same cooperative capabilities, observability and proactivity, that humans use. We show how agents can use observation of the environment and of teammates' actions to estimate the teammates' beliefs without generating unnecessary messages; we also show how agents can anticipate information needs among the team members and proactively communicate the information, reducing the total volume of communication. Finally, we present several experiments that validate the system developed, explore the effectiveness of different aspects of observability and introduce the scalability of the use of observability with respect to the number of agents in a system.

Every team matters: The contribution educational psychology can make to effective teamwork

Teamwork has never been more topical. This is evident not only in the practice of educational psychologists but also in the dominant discourse surrounding government policy. Multi-agency collaboration lies at the heart ofEvery Child Matters. The underlying assumption is that multi-agency teamwork is the best way to work, ensuring the most complete and effective service to children and their families. While not disputing this, we feel that the complexities of teamwork are not sufficiently considered. This paper will address some of the central issues related to the psychology of teams from the experience of three educational psychologists. We consider the complexities of teamwork by drawing on psychodynamic, systemic and social constructionist thinking. A grounded understanding and practical application of the psychology of teams is essential to effective practice. We support this with some reference to our own varied experience of work with multi-agency teams. This understanding will be used to suggest some ways to develop effective teamwork. Our focus throughout will be on adult, professional groups: the dynamics, group processes and our psychological understanding of them. Our intention is to provoke debate and to provide a ‘real’ account of the experience of three educational psychologists, all of whom have worked for a number of years within teams and independently of teams.

‘When is a teacher not a teacher?’: knowledge creation and the professional identity of teachers within multi-agency teams

Education is centre stage in current UK government initiatives to promote multi-agency team work. This paper draws on a research project which explored the way in which multi-disciplinary teams work and learn together in their practice with children, to consider the implications of ‘joined-up’ practice for theorizing dilemmas of knowledge creation and identity transformation for professionals in multi-agency teams. The paper focuses primarily on the experiences of education professionals. We exemplify some dilemmas of ‘joined-up’ team participation in specific workplace activities involving knowledge exchange. We then explore the impact of belonging to multi-agency teams on professional roles, identities and learning. The paper then summarizes strategies which professionals used for resolving dilemmas around learning and knowledge creation, and considers how participating in shared workplace activities might enable or constrain professionals to consolidate their professional identities and learning. Drawing on theoretical research into workplace participation and professional learning, the paper examines implications for theorizing the professional identity of teachers in multi-agency team work, within a systemic model that takes account of: creating new knowledge and practice; enhancing professional identity; and building inter-professional communities.

Investigating the impact of working in multi-agency service delivery settings in the uk on early years practitioners’ beliefs and practices

In the UK Centres of Excellence were funded by the DfES to model high quality, multi-agency, early years services for young children and their families. They were precursors to Children’s Centres to be established across the UK. Early Excellence Centres were evaluated at national and local levels. This article will draw on data from local evaluations of two contrasting Centres of Excellence: interviews with staff to explore shifts in their perspectives on the rhetoric and reality of multi-agency team work; and responses by staff to vignettes representing conflicts and dilemmas in the delivery of services. The article explores methodological complexities of accessing values underpinning action in practice in multi-agency teams working in inner cities.

Health professionals in multi-disciplinary and multi-agency teams: Changing professional practice

The article draws on an Economic and Social Research Council (ESRC)-funded research project that aimed to investigate the reality behind the rhetoric of “joined up thinking”. The research project was a qualitative, multi-method study involving three phases, including observation and documentary analysis; interviews; and focus groups around decision making and knowledge sharing. The article reflects on the perspectives and experiences of health professionals and their colleagues in multi-agency teams about the impact of multi-agency teamwork on their professional knowledge and learning, and on their ways of working. Actual and potential conflicts between professionals are explored about models of understanding, about roles, identities, status and power, about information sharing, and around links with other agencies. Dilemmas of team building and of conflicting values and knowledge are exemplified from health professionals' accounts, using theoretical models of “communities of practice” and “activity theory”. The article presents groups of strategies that health professionals and their colleagues in multi-agency, multi-professional teams use to overcome barriers and to strengthen team cohesion. The conclusion reflects on some implications of our findings in theory and practice for professionalism within integrated, multi-professional teams that are building new ways of working.

Automated Assistants for Analyzing Team Behaviors

Multi-agent teamwork is critical in a large number of agent applications, including training, education, virtual enterprises and collective robotics. The complex interactions of agents in a team as well as with other agents make it extremely difficult for human developers to understand and analyze agent-team behavior. It has thus become increasingly important to develop tools that can help humans analyze, evaluate, and understand team behaviors. However, the problem of automated team analysis is largely unaddressed in previous work. In this article, we identify several key constraints faced by team analysts. Most fundamentally, multiple types of models of team behavior are necessary to analyze different granularities of team events, including agent actions, interactions, and global performance. In addition, effective ways of presenting the analysis to humans is critical and the presentation techniques depend on the model being presented. Finally, analysis should be independent of underlying team architecture and implementation. We also demonstrate an approach to addressing these constraints by building an automated team analyst called ISAAC for post-hoc, off-line agent-team analysis. ISAAC acquires multiple, heterogeneous team models via machine learning over teams' external behavior traces, where the specific learning techniques are tailored to the particular model learned. Additionally, ISAAC employs multiple presentation techniques that can aid human understanding of the analyses. ISAAC also provides feedback on team improvement in two novel ways: (i) It supports principled “what-if” reasoning about possible agent improvements; (ii) It allows the user to compare different teams based on their patterns of interactions. This paper presents ISAAC's general conceptual framework, motivating its design, as well as its concrete application in two domains: (i) RoboCup Soccer; (ii) software agent teams participating in a simulated evacuation scenario. In the RoboCup domain, ISAAC was used prior to and during the RoboCup '99 tournament, and was awarded the RoboCup Scientific Challenge Award. In the evacuation domain, ISAAC was used to analyze patterns of message exchanges among software agents, illustrating the generality of ISAAC's techniques. We present detailed algorithms and experimental results from ISAAC's application.

The Communicative Multiagent Team Decision Problem: Analyzing Teamwork Theories and Models

Despite the significant progress in multiagent teamwork, existing research does not address the optimality of its prescriptions nor the complexity of the teamwork problem. Without a characterization of the optimality-complexity tradeoffs, it is impossible to determine whether the assumptions and approximations made by a particular theory gain enough efficiency to justify the losses in overall performance. To provide a tool for use by multiagent researchers in evaluating this tradeoff, we present a unified framework, the COMmunicative Multiagent Team Decision Problem (COM-MTDP). The COM-MTDP model combines and extends existing multiagent theories, such as decentralized partially observable Markov decision processes and economic team theory. In addition to their generality of representation, COM-MTDPs also support the analysis of both the optimality of team performance and the computational complexity of the agents' decision problem. In analyzing complexity, we present a breakdown of the computational complexity of constructing optimal teams under various classes of problem domains, along the dimensions of observability and communication cost. In analyzing optimality, we exploit the COM-MTDP's ability to encode existing teamwork theories and models to encode two instantiations of joint intentions theory taken from the literature. Furthermore, the COM-MTDP model provides a basis for the development of novel team coordination algorithms. We derive a domain-independent criterion for optimal communication and provide a comparative analysis of the two joint intentions instantiations with respect to this optimal policy. We have implemented a reusable, domain-independent software package based on COM-MTDPs to analyze teamwork coordination strategies, and we demonstrate its use by encoding and evaluating the two joint intentions strategies within an example domain.

The communicative multiagent team decision problem

Despite the significant progress in multiagent teamwork, existing research does not address the optimality of its prescriptions nor the complexity of the teamwork problem. Without a characterizatio...

RoboCup rescue: A grand challenge for multiagent and intelligent systems

Disaster rescue is one of the most serious social issues that involves very large numbers of heterogeneous agents in the hostile environment. The intention of the RoboCup Rescue project is to promote research and development in this socially significant domain at various levels, involving multiagent teamwork coordination, physical agents for search and rescue, information infrastructures, personal digital assistants, a standard simulator and decision-support systems, evaluation benchmarks for rescue strategies, and robotic systems that are all integrated into a comprehensive system in the future. For this effort, which was built on the success of the RoboCup Soccer project, we will provide forums of technical discussions and competitive evaluations for researchers and practitioners. Although the rescue domain is intuitively appealing as a large-scale multiagent and intelligent system domain, analysis has not yet revealed its domain characteristics. The first research evaluation meeting will be held at RoboCup-2001, in conjunction with the Seventeenth International Joint Conference on Artificial Intelligence (IJCAI-2001), as part of the RoboCup Rescue Simulation League and RoboCup/AAAI Rescue Robot Competition. In this article, we present a detailed analysis of the task domain and elucidate characteristics necessary for multiagent and intelligent systems for this domain. Then, we present an overview of the RoboCup Rescue project.

Agent Assistants for Team Analysis

With the growing importance of multiagent team-work, tools that can help humans analyze, evaluate, and understand team behaviors are also becoming increasingly important. To this end, we are creating isaac, a team analyst agent for post hoc, offline agent-team analysis. ISAAC'S novelty stems from a key design constraint that arises in team analysis: Multiple types of models of team behavior are necessary to analyze different granularities of team events, including agent actions, interactions, and global performance. These heterogeneous team models are automatically acquired by machine learning over teams' external behavior traces, where the specific learning techniques are tailored to the particular model learned. Additionally, ISAAC uses multiple presentation techniques that can aid human understanding of the analyses. This article presents ISAAC'S general conceptual framework and its application in the RoboCup soccer domain, where ISAAC was awarded the RoboCup Scientific Challenge Award.