10th International Workshop on Engineering Multi-Agent Systems
9-10 May 2022, Auckland, New Zealand


Multiagent systems (MAS)—systems composed of autonomous agents who interact with each other— are ideally suited to realizing modern software applications. For instance, MAS ideas are well-suited to the IoT, and applications that support engagements between humans and organizations in domains such as business, health, and finance. Moreover, with the advent of microservices, we see a clear shift in the software industry from monolithic applications to applications constituted from autonomous, interactive services—in essence, agents.

The EMAS workshop is a forum for presenting and discussing original ideas and work on software abstractions, methodologies, languages, and tools for engineering loosely-coupled, adaptive, high-performance, and scalable MAS.

As AAMAS 2022 will be a virtual event, EMAS 2022 will adopt a virtual format similar to that used for EMAS 2021.

Schedule

Session 1: Invited Talk [Chair: Amit K. Chopra]

Simon Miles. Testing agent-oriented systems for emergent behaviour 🎥

Emergence is a key reason why testing agent-oriented software can be challenging. The behaviour of the system over time due to cumulative individual interactions of diverse agents in a dynamic environment can be unpredictable in nature and scale. Different approaches are taken to this problem from trying to enforce certain guarantees to recovering at run-time when the system acts in contradiction to design goals. The nature of emergence means we need to consider how to engineer the system so that the emerging behaviour can be apparent to tests and/or detected by agents, and so part of the challenge is testing that this infrastructure is itself functioning. It is valuable to consider the overall space of this problem and how to give engineers consistent ways to analyse emergence from the perspective of testing a system. This includes considerations such as that unpredicted emergent behaviour is not always detrimental but can be beneficial, arguably a reason for using an agent-oriented design approach in the first place, and that a system could include agents whose interests may not be served by revealing when novel behaviour emerges, e.g. in competitive markets. In this talk, I will discuss these challenges and ideas for addressing some of them.

Simon Miles is a Reader in Computer Science at King's College London, UK. He conducts research in the areas of agent-based systems, particularly normative systems, agent-based modelling and agent-oriented software engineering, and data provenance. He has published widely in these areas, co-authored a W3C standard on provenance (PROV), and led research initiatives including the multi-disciplinary Centre for Urban Science and Progress London (CUSP London) and the Agents & Intelligent Systems group at King's. He was previously a researcher at the University of Southampton after graduating from his PhD at University of Warwick.

Session 2: Agent-Oriented Software Engineering [Chair: Rafael Bordini]

Patrick Gavigan and Babak Esfandiari. Quantifying the Relationship Between Software Design Principles and Performance in Jason 🎥 📖

We investigated the relationship between various design approaches of AgentSpeak code for Jason Beliefs-Desires-Intentions (BDI) agents and their performance in a simulated automotive collision avoidance scenario. Also explored was how the approaches affected software maintainability, assessed through coupling, cohesion, and cyclomatic complexity. We then compared each agent's performance, specifically their reasoning cycle duration and their responsiveness. Our findings revealed that agents with looser coupling and higher cohesion are more responsive to stimuli, implying that more maintainable AgentSpeak can result in better performing agents. Performance was inversely related to cyclomatic complexity.

Tobias Ahlbrecht. An algorithmic debugging approach for BDI agents 🎥 📖

Debugging agent systems can be rather difficult. It is often noted as one of the most time-consuming tasks during the development of cognitive agents. Algorithmic (or declarative) debugging is a semi- automatic technique, where the developer is asked questions by the de- bugger in order to locate the source of an error. We present how this can be applied in the context of a BDI agent language, demonstrate how it can speed up or simplify the debugging process and reflect on its advantages and limitations.

Samuele Burattini, Angelo Croatti, Alessandro Ricci, Andrei Ciortea, Danai Vachtsevanou, Jeremy Lemee and Simon Mayer. Agent-Oriented Visual Programming for the Web of Things 🎥 📖

In this paper we introduce and discuss an approach for multi-agent-oriented visual programming. This aims at enabling individuals without programming experience but with knowledge in specific target domains to design and (re)configure autonomous software. We argue that, compared to procedural programming, it should be simpler for users to create programs when agent abstractions are employed. The underlying rationale is that these abstractions, and specifically the belief-desire-intention architecture that is aligned with human practical reasoning, match more closely with people's everyday experience in interacting with other agents and artifacts in the real world. On top of this, we designed and implemented a visual programming system for agents that hides the technicalities of agent-oriented programming using a blocks-based visual development environment that is built on the JaCaMo platform. To further validate the proposed solution, we integrate the Web of Things (WoT) to let users create autonomous behaviour on top of physical mashups of devices, following the trends in industrial end-user programming. Finally, we report on a pilot user study where we verified that novice users are indeed able to make use of this development environment to create multi-agent systems to solve simple automation tasks.

Yi Yang and Tom Holvoet. Making Model Checking Feasible for GOAL 🎥 📖

Agent Programming Languages have been studied for over 20 years for programming complex decision-making for autonomous sys- tems. The GOAL agent programming language is particularly interesting as it does not require any preprogrammed planning by developers, but instead relies on automated planning based on beliefs and goals to de- termine its behavior. Model checking is a powerful verification technique to guarantee the safety of an autonomous system. Despite studies of model checking in other agent programming languages, GOAL lacks support for model checking of GOAL programs. The fundamental challenge is to make GOAL programs feasible for model checking in the first place. In this paper, we tackle this fundamental issue. We devise an algorithm for transforming a (considerable) subset of GOAL programs to a tran- sition system that is equivalent in terms of operational semantics, en- abling model checking. We prove the correctness of this algorithm. We implement the transformation algorithm, and we discuss the scalability through Blocks World examples of increasing size. Moreover, we point out that we will extend the applicability of the transformation algorithm and its implementation to all stratified GOAL programs.

Session 3: Abstractions and Agent-Based Modeling [Chair: Nadin Kökciyan]

Jérémy Lemée, Danai Vachtsevanou, Simon Mayer and Andrei Ciortea. Signifiers for Affordance-driven Multi-Agent Systems 🎥 📖

The ecological psychologist James J. Gibson defined the notion of affordances to refer to what action possibilities environments offer to animals. In this paper, we show how (artificial) agents can discover and use affordances in a Multi-Agent System (MAS) environment to achieve their goals. To indicate to agents what affordances are present in their environment and whether it is likely that these may help the agents to achieve their objectives, the environment may expose signifiers while taking into account the current situation of the environment and of the agent. On this basis, we define a Signifier Exposure Mechanism that is used by the environment to compute which signifiers should be exposed to agents in order to permit agents to only perceive signifiers that are likely to be relevant to them, and thereby increase their efficiency. If this is successful, agents can interact with partially observable environments more efficiently because the signifiers indicate the affordances they can use towards which purposes. Signifiers thereby facilitate the exploration and the exploitation of MAS environments. An implementation of signifiers and of a Signifier Exposure Mechanism is presented within the context of a Hypermedia Multi-Agent System and the utility and efficiency of this model is presented through the development of a scenario.

Antonio Carlos Rocha Costa. A Short Note on the Bounds of the Organizational Approach to MAS 🎥 📖

This short note claims that societal architectures, not just organizational architectures, are the appropriate architectural forms for framing the conception, design, and implementation of full-fledged multiagent systems, that is, MAS that are able to computationally model all the essential characteristics of general social systems.

Stefano Mariani, Marco Picone and Alessandro Ricci. About Digital Twins, agents, and multiagent systems: a cross-fertilisation journey 🎥 📖

Digital Twins (DTs) are rapidly emerging as a fundamental brick of engineering cyber-physical systems, but their notion is still mostly bound to specific business domains (e.g. manufacturing), goals (e.g. product design), or application domains (e.g. the Internet of Things). As such, their value as general purpose engineering abstractions is yet to be fully revealed. In this paper, we relate DTs with agents and multiagent systems, as the latter are arguably the most rich abstractions available for the engineering of complex socio-technical and cyber-physical systems, and the former could both fill in some gaps in agent-oriented engineering and benefit from an agent-oriented interpretation—in a cross-fertilisation journey.

Hussein Marah and Moharram Challenger. Intelligent Agents and Multi Agent Systems for Modeling Smart Digital Twins 🎥 📖

Recent years witnessed a significant digital transformation in most aspects of life and various disciplines. A new technology concept named a Digital Twin has emerged. Digital Twin is mainly utilized to create a virtual representation (digital assets) of physical components (physical assets) which are connected and synchronized with each other. Digital Twin could be used in multiple tasks like diagnosing, forecasting, and visualization, decreasing the cost of designing, implementing, and using the physical assets. Nevertheless, because of the complexity level encountered when merging both cyber and physical parts, building a Digital Twin is considered a quite challenging job, as many requirements should be addressed during the design and the implementation processes. Thus, it's crucial to choose a reliable approach that can implement such a complex system. Therefore, this paper aims to introduce an intelligent approach driven by using the Multi-agent System architecture that could reduce the level of complexity while building a Digital Twin. Intelligent agents enabled us to deal with the challenges imposed by building CPS as well as the integration with the Digital Twin. In this approach, different components of CPS are represented as autonomous agents in the Digital Twin environment. Every individual agent represents a specific part of the system and acts intelligently and cooperatively with other agents to achieve the system's objectives.

Panel [Chair: Michael Winikoff]

10 Years of EMAS: Looking Backwards, Looking Forwards 🎥

Panelists:

  • Prof Jorge Gómez-Sanz (Complutense University of Madrid)
  • Prof Alessandro Ricci (University of Bologna)
  • Prof Pınar Yolum (Utrecht University)
Discussion:
  • How has EMAS research changed over the past decade? What has remained constant?
  • What is the biggest (non-research) challenge for the EMAS community?
  • What is the future of EMAS?

Session 4: Reasoning and Negotiation [Chair: Pradeep Murukannaiah]

Barbara Dunin-Kęplicz and Andrzej Szałas. Modeling and Shadowing Paraconsistent BDI Agents 🎥 📖

The BDI model of rational agency has been studied for over three decades. Many robust multiagent systems have been developed, and a~number of BDI logics have been studied. Following this intensive development phase, the importance of integrating BDI models with inconsistency handling and revision theory have been emphasized. There is also a~demand for a tighter connection between BDI-based implementations and BDI logics. We address these postulates by introducing a novel, paraconsistent logical BDI model close to implementation, with building blocks that can be represented as SQL/rule-based databases. Importantly, tractability is achieved by reasoning as querying. This stands in a sharp contrast to the high complexity of known BDI logics. We also extend belief shadowing, a shallow and lightweight alternative to deep and computationally demanding belief revision, to encompass agents' motivational attitudes.

Phillip Sloan and Nirav Ajmeri. Commitment-Based Negotiation Semantics 🎥 📖

Negotiation is a key form of interaction in multi-agent sys- tems. Negotiation enables agents to come to a mutual agreement about some goal or plan of action. Current negotiation approaches use tradi- tional interaction protocols which do not capture the normative meaning of interactions and often restrict agent autonomy. These traditional nego- tiation approaches also have difficulty capturing accountability — a key requirement for creating an ethical agent. This paper seeks to address this gap in maintaining autonomy and establishing accountability re- quirements during negotiation. We propose Nala, a commitment-based negotiation semantics. Nala uses commitments to help provide norma- tive meaning to agent interactions. The nature of commitments establish accountability requirements between agents in negotiation. We illustrate Nala's usage via a case study using a game scenario where agents partic- ipate in negotiation to bring about their goals in a research constrained environment.

Thimjo Koça, Catholijn Jonker and Tim Baarslag. Enabling Negotiating Agents to Explore Very Large Outcome Spaces 🎥 📖

This work presents BIDS (Bidding using Diversified Search), an algorithm that can be used by negotiating agents to search very large outcome spaces. BIDS provides a balance between being rapid, accurate, diverse, and scalable search, allowing agents to search spaces with as many as 10250 possible outcomes on very run-of-the-mill hardware. We show that our algorithm can be used to respond to the three most common search queries employed by 87% of all agents from the Automated Negotiating Agents Competition. Furthermore, we validate one of our techniques by integrating it into negotiation platform GeniusWeb, to en- able existing state-of-the-art agents (and future agents) to scale their use to very large outcome spaces.

Session 5: Frameworks and Applications [Chair: Andrei Ciortea]

Andreas Brännström and Juan Carlos Nieves. A Framework for Developing Interactive Intelligent Systems in Unity 🎥 📖

This paper introduces a lightweight framework for implementing intelligent interactive systems (IIS). In particular, systems that integrate symbolic knowledge bases for reasoning, planning and rational decision-making in interactions with humans. This is done by integrating Web Ontology Language (OWL)-based reasoning and Answer Set Programming (ASP)-based planning software. In order to provide interactive user components, the framework is encompassed in a widely used game development tool, Unity. The proposed framework, UnityIIS, is the first approach for integrating OWL together with ASP in Unity. Its central functionalities for knowledge representation and knowledge revision is presented together with an example application created in the framework. A chatbot agent, embodied in Augmented Reality, is designed, following a Belief, Desire, Intention (BDI) agent model. The set of tools that the framework provides can be applied for developing IIS research prototypes as well as being an asset in teaching practices.

Sanchayan Bhunia, Angelo Ferrando, Viviana Mascardi and Chiara Vitale. MAiS: exploiting JADE as a Multi-Agent simulator of the Immune System 🎥 📖

The immune system is the second most complex biological system after the brain. It consists in millions of cells, of various nature, in- teracting amongst them to keep the organism safe from external enemies (pathogens), such as viruses and bacteria. To better understand how the immune system works, and how it reacts to certain diseases and cures, simulations have been proposed over the years. Amongst them, we may find agent-based ones where the organism's actors, like cells, antibodies, viruses, and so on, are represented as agents. In this paper, we present the initial design and development of an agent-based simulation of the immune system using a well-known agent framework, JADE. We present the engineering choices we made and the instantiation of some steps of the secondary immune system response. We discuss the implementation in JADE, and we present some experimental results.

Tabajara Krausburg, Rafael H.Bordini and Jürgen Dix. Modelling a Chain of Command in the Incident Command System using Sequential CFGs 🎥 📖

Disaster response is a major challenge given the social and economic impact on the communities affected by disaster incidents. We investigate how coalition formation can be used for the problem of forming a hierarchy of resources (e.g., personnel responding to the incident). As a case study, we consider the roaring river flood scenario and model the Incident Command System (ICS) framework—providing guidelines on cooperatively responding to disaster incidents. Our main tool is sequential characteristic-function games induced by size-based valuation structures. We show that our approach can deliver a hierarchy required by the Operations Section in the ICS and provides a promising way to analyse the computed solution. Future work will focus on collecting feedback from experts on the topic regarding the proposed modelling.

Samuel Christie and Amit Chopra. Bruno: Garbage-Collecting Business Information 🎥 📖

Business communications contain information of value to the parties involved. Some information represents and enables currently on- going transactions. For purposes of efficiency, parties would want such information to be available in local storage, which is fast to access but possibly expensive per unit capacity. Other information is usually con- signed to “offline” archival and may support a variety of purposes, e.g., analysis, reporting, meeting legal requirements, and so on. Archival storage is typically both significantly slower and less expensive compared to local storage. Businesses typically have policies based on which they consign information to archive. A related challenge is how to understand and process business events whose context has been archived. We refer to the idea of archiving information as garbage collection. We present a multiagent conception of garbage collection based on proto- cols. Agents represent the involved parties and enact protocols. Protocol enactments represent transactions. An agent applies policies to decide which transactions to archive. Although the policies can be arbitrary, we show how to exploit the protocol to inform them. We also discuss how an agent may handle events whose context may be in archive.

Accepted Papers

  • Antonio Carlos Rocha Costa. A Short Note on the Bounds of the Organizational Approach to MAS
  • Hussein Marah and Moharram Challenger. Intelligent Agents and Multi Agent Systems for Modeling Smart Digital Twins
  • Tabajara Krausburg, Rafael H. Bordini and Jürgen Dix. Modelling a Chain of Command in the Incident Command System using Sequential CFGs
  • Sanchayan Bhunia, Angelo Ferrando, Viviana Mascardi and Chiara Vitale. MAiS: exploiting JADE as a Multi-Agent simulator of the Immune System
  • Patrick Gavigan and Babak Esfandiari. Quantifying the Relationship Between Software Design Principles and Performance in Jason
  • Phillip Sloan and Nirav Ajmeri. Commitment-Based Negotiation Semantics
  • Andreas Brännström and Juan Carlos Nieves. A Framework for Developing Interactive Intelligent Systems in Unity
  • Barbara Dunin-Kęplicz and Andrzej Szałas. Modeling and Shadowing Paraconsistent BDI Agents
  • Tobias Ahlbrecht. An algorithmic debugging approach for BDI agents
  • Yi Yang and Tom Holvoet. Making Model Checking Feasible for GOAL
  • Stefano Mariani, Marco Picone and Alessandro Ricci. About Digital Twins, agents, and multiagent systems: a cross-fertilisation journey
  • Thimjo Koca, Catholijn Jonker and Tim Baarslag. Enabling Negotiating Agents to Explore Very Large Outcome Spaces
  • Jérémy Lemée, Danai Vachtsevanou, Simon Mayer and Andrei Ciortea. Signifiers for Affordance-driven Multi-Agent Systems
  • Samuele Burattini, Angelo Croatti, Alessandro Ricci, Andrei Ciortea, Danai Vachtsevanou, Jeremy Lemee and Simon Mayer. Agent-Oriented Visual Programming for the Web of Things
  • Samuel Christie and Amit Chopra. Bruno: Garbage-Collecting Business Information

Important Dates

Submission deadline March 11, 2022 March 4, 2022
Author notification April 8, 2022
Camera-ready deadline April 29, 2022 April 22, 2022
EMAS May 9-10, 2022

Information for Authors

Final Paper & Video Submission

  • Produce the final version of your paper (up to 20 pages), taking the reviews into account.
  • Upload your updated paper together with the video presentation (see below) by April 29th 2022.

Conference Presentation

  • The conference presentations will take place via Microsoft Teams.
    • You can participate with your browser or install the Teams application.
    • The invitation will be distributed to everyone who has registered.
  • We have grouped all presentations into sessions, most with 4 papers (one with 3). We assume that most will have watched the videos before the presentation, so a presentation should not try to replicate the video, but rather highlight the main achievements (without details).
  • As we want ample time for discussion, we ask you to finish your presentation within at most 8-10 minutes (very strict deadline). Very important questions (understanding notation or so) can then be asked (if necessary). This permits us to have 20 minutes for discussion after the presentations.

Video Submission

  • To foster in-depth discussion, authors are asked to prepare a video (15 minutes maximum) that will be shared with EMAS attendees before the start of the workshop.
  • Videos will be uploaded to YouTube.
    • Please make sure that you own all the copyrights of the material in the video.
  • To send us your video, please provide it via Cryptshare to t...@t...l.de (up to 20GB).
    • Submit the abstract and your short bio for the video description (together with the video file or to the same e-mail address).
    • Also include the final version of your paper (see above).
Papers and links to videos will be made available through the EMAS website.

Committees

Organising Committee

Programme Committee

  • Natasha Alechina, Utrecht University
  • Luciano Baresi, Politecnico di Milano
  • Cristina Baroglio, Università di Torino
  • Rafael Bordini, PUCRS
  • Daniela Briola, University of Insubria
  • Rafael C. Cardoso, University of Aberdeen
  • Moharram Challenger, University of Antwerp
  • Amit Chopra, Lancaster University
  • Andrei Ciortea, University of St. Gallen
  • Rem Collier, UCD
  • Stefania Costantini, Univ. dell'Aquila
  • Mehdi Dastani, Utrecht University
  • Maiquel de Brito, Federal University of Santa Catarina
  • Davide Dell'Anna, Delft University of Technology
  • Louise Dennis, University of Manchester
  • Juergen Dix, Clausthal University of Technology
  • Angelo Ferrando, University of Genova
  • Lars-Ake Fredlund, Universidad Politécnica de Madrid
  • Stéphane Galland, UBFC - UTBM
  • Jorge Gomez-Sanz, Universidad Complutense de Madrid
  • Zahia Guessoum, Université de Paris 6 and Université de Reims Champagne Ardenne
  • James Harland, RMIT University
  • Vincent Hilaire, UTBM/IRTES-SET
  • Tom Holvoet, Katholieke Universiteit Leuven
  • Jomi Fred Hubner, Federal University of Santa Catarina
  • Yves Lespérance, York University
  • Viviana Mascardi, University of Genova
  • Philippe Mathieu, University of Lille
  • John-Jules Meyer, Utrecht University
  • Roberto Micalizio, Universita' di Torino
  • Jörg P. Müller, TU Clausthal
  • Alessandro Ricci, University of Bologna
  • Luca Sabatucci, ICAR-CNR
  • Jaime Sichman, University of São Paulo
  • Viviane Torres da Silva, IBM
  • Shihan Wang, Utrecht University
  • Gerhard Weiss, University Maastricht
  • Michael Winikoff, Victoria University of Wellington
  • Neil Yorke-Smith, Delft University of Technology
  • Rym Zalila-Wenkstern, The University of Texas at Dallas

Steering Committee

  • Matteo Baldoni
  • Rafael Bordini
  • Mehdi Dastani
  • Jürgen Dix
  • Amal El Fallah Seghrouchni
  • Brian Logan
  • Jörg P. Müller
  • Alessandro Ricci
  • Danny Weyns
  • Michael Winikoff
  • Rym Zalila-Wenkstern