SlideShare ist ein Scribd-Unternehmen logo
1 von 70
Downloaden Sie, um offline zu lesen
LECTURE 10:
Cooperation in MAS (IV):
implicit methods
    Artificial Intelligence II – Multi-Agent Systems
        Introduction to Multi-Agent Systems
              URV, Winter-Spring 2010
Outline of the lecture

   Implicit cooperation in MAS
     Indirect cooperation through the
     environment
     Societal views of MAS
       Electronic institutions
       Organizational structures
Coordination [recall past lectures]
An activity is a set of potential operations an actor
(an agent playing a certain role) can perform, with
the aim of achieving a given goal or set of goals
Coordination could be defined as the process of
managing dependencies between activities. By
such process an agent reasons about its local
actions and the foreseen actions that other
agents may perform, with the aim to make the
community to behave globally in a coherent
manner
Cooperation hierarchy [last lectures]
                                     MAS


                    Independent                  Cooperative
                   Self-interested               Benevolent



        Discrete         Emergent              With               Without
                                           communication -     communication -
                                              Explicit            Implicit


           Reactive
           systems
                                 Deliberative         Negotiators

Partial Global                                                          Auctions
Planning
                       Coalition                   Voting
                       formation                                    Contract Net
Implicit cooperation
 A group of distributed cooperative agents
 behaves in a socially coordinated way in the
 resolution of a global problem without an
 explicit exchange of communication
 messages
 In many cases the environment acts as the
 (indirect) interaction mechanism
Motivation (I)
 Cases in which explicit coordination cannot be
 applied:
   Speed: it takes too long to communicate with others – by
   then the opportunities are missed
      E.g. Football game – simple signals may work, but
      lengthy explanations don't...
      In general, very dynamic environments
   Security: not wanting others to know
   what your plans are
Motivation (II)

   Complexity: some agents may be too simple to
   deal with the complexity of generating and
   understanding complex plans
     Reactive rule-based robots
     Complexity of Partial Global Planning or coalition
     formation
   Lack of a communication channel: there may
   actually be no way to communicate
     Physical robots with limited communication range
Options for implicit cooperation

   Observe the behaviour of the other
   agents, and react accordingly
     Indirect cooperation through the effects on
     the environment of the actions of each
     agent
   Imposing a structure on the MAS
Emergent Coordination [recall past lectures]
  Coordination in cases where:
    There is no communication between agents
    There is no mechanism for enforcing a-priori
    social rules / laws
    Agents have their own agenda/goals
  The resulting coordination is emergent
  and cannot be said to be based on joint
  plans or intentions
Basic difference
  Emergent coordination: agents are self-
  interested, they do not care about the other
  agents in the system, there isn’t any high
  level design of the emergent behaviour
  Implicit coordination (also giving rise to
  emergent coordinated global behaviour):
  although agents do not communicate with
  each other, the designer of the system
  intends to provoke the emergence of the
  socially intelligent problem solving activities
Implicit coordination example:
Network Routing
  Network Routing problems
  are challenging. Solutions
  need to be:
     Dynamic
     Robust
  Network of N nodes, L links.
  Traffic flows as packets
  traverse the network
  There are protocols that
  compute cumulative shortest
  path measures
Ants discover shortest paths
Network Ants
 Ants randomly explore
 the network until they
 find a specific node
    They mark the traversed
    paths with “pheromone”
 Ants seeking
 destinations follow
 pheromone trails
 Pheromones degrade
 over time
                                      Robust
                                       Stable
                              Gradual Change
Pheromone tables
 Each node contains a table of probabilities
 (pheromone table) for each possible
 destination in the network
   In a 30-nodes network, each node keeps 29
   tables
 The entries on the tables are the probabilities
 which influence the ants’ selection of the next
 node on the way to their destination node
 Pheromone laying = updating probabilities
Pheromone tables example
 A network with 6 nodes,
 node 1 is connected with
 nodes 2, 4 and 5.                                Next node


 The pheromone tables in
                                                    2          4      5
 node 1 would look like this:
 For instance, if an ant                      2    0.90       0.05   0.05

 arrives at node 1 and wants
 to go to node 3, the most                    3    0.25       0.60   0.15

 probable route is through      Destination   4    0.10       0.85   0.05

 node 4 (but it may also        node          5    0.10       0.10   0.80


 decide to go through nodes
 2 or 5)                                      6    0.40       0.30   0.30
Simulation (I)

 At each step, ants can be launched from any
 node in the network, with a random
 destination node
 Ants move from node to node, selecting the
 next node to move to according to the
 probabilities in the pheromone tables for their
 destination node
   Pheromone tables are initialized with random
   values
Simulation (II)
  When ants arrive at a node, they update the
  probabilities of that node’s pheromone table
  entries corresponding to their source node
  They alter the table to increase the probability
  pointing to their previous node
  Ants moving away from their source node
  can only directly affect those ants for which it
  is the destination node
Pheromone laying example
 An ant has to go from node 3 to node 2; in
 the way, it travels from node 4 to node 1
  First, it modifies the table in node 1 corresponding
  to node 3, increasing the probability of selecting
  the link to node 4
  After that, it selects the next node randomly
  according to the probabilities of the table in node
  1 corresponding to node 2

      3    …      4        1        …     2
Increasing/decreasing pheromones

 Pheromones are increased with the following
 formula
          p = (p_old + Δ(p)) / (1 + Δ(p))
 As all the entries must add up to 1, the other
 entries have to be decreased as follows
          p = p_old / (1 + Δ(p))
 Note that probabilities may never be 0
Basic ideas of implicit cooperation

  Agents do not talk to each other directly
  Agents can modify the environment, and
  these modifications influence the behaviour
  of the other agents in the system
  All the agents contribute towards a useful
  global behaviour of the community
Reasoning mechanisms for coordination

 Thinking about individual agents
   Methods that allow building a model of the
   other agents of the system
 Thinking about the whole agent’s society
   Methods that try to impose some kind of
   rules/laws/structure/organisation in the multi-
   agent system
Agent Modelling (I)
  Even if you cannot talk to the other
  agents you may still want to reason about
  them
  Main methods:
    Recursive Modelling Methods
      Assume the others have a similar structure to you
      – and may have a model of you...
      Try to deduce their beliefs/desires/intentions from
      their actions on the environment
Agent modelling (II)

   Plan Recognition
     Analyse the sequences of activities of other agents
     and try to discover their plans (and, from them,
     identify the potential end goals of their actual
     actions)
   Game Playing / Game Tree Search:
     Modelling opponents
     For example, using minimax search
     [Recall Game Theory in Artificial Intelligence]
Thinking about Society
Common approaches include:
  Social Laws: global rules which agents follow and
  lead to “coherent behaviour”, either instilled in the
  agent or communicated when entering the
  environment (e.g. - “driving on the right hand side”)
  Social Power Relations: a theory of dependence
  relations, in particular to model goal adoption (e.g.
  carrying out work on behalf of a superior)
  Electronic Institutions
  Organizational structures
Institutions as Social Structures

  Social Structures define a social level to
  enhance coordination by means of
  interaction patterns
  Institutions are a kind of social structure
  where a corpora of constraints shape the
  behaviour of the members of a group
Institution components

 The definition of a (human) Institution
 usually includes:
   Norms about the interactions
   Conventions: acceptable (and unacceptable)
   actions within the institution
   Procedures and protocols to be followed
e-Institutions
   An e-Institution is the computational model
   of an institution through
    The specification of the institution’s norms in
    some suitable formalism
    The formal specification of the institution’s
    admissible procedures and protocols, which
    follow the established conventions
E-Institutions and MAS
 In the context of MAS, e-institutions:
   reduce uncertainty of other agents’ behaviour
   reduce misunderstanding in interaction
   allow agents to foresee the outcome of an
   interaction
   simplify the decision making process (by reducing
   the possible actions)

   Agent behaviour guided by Norms
Why a Language for Norms?
                            Laws,
                            Laws,                    [Natural Language]
                         regulations
                          regulations
too abstract and
     vague

                       Language for norms
                       Language for norms            [Formal Language]
more concrete      (Formal & Computational)
                    (Formal & Computational)


                                          Electronic Institutions
       Normative Agents

           Norms in                            Norm enforcement
          deliberation
              cycle                              mechanisms
Influence of norms in the BDI deliberation cycle
                                                      input


                             Agent        sensors
                                  perception                   E
        state
                                    How is the                 N
                                    world now?                 V
                                                               I
                                  What if I perform
                                                               R
                                                               O
        KB                          action A?
                                                               N
                                                               M
                                  Which action do              E
                                    I choose?                  N
                                                               T
        goals
                                          actuators
                  norms
                 (obligations,
                permissions...)                       action
AMELI (I)
  AMELI is an institution middleware that is based
  in a formal electronic institution specification tool
  (ISLANDER), developed at IIIA
  The ISLANDER framework is composed of:
    A Dialogical Framework
       Linguistic and social structure (roles) to give meaning to
       agent interactions, communication language
    A Performative Structure
       scenes and      relationships   between    scenes    (e.g.
       precedence)
    Rules
       Conventions to be followed, social commitments
AMELI (II)

  Two hypotheses:
   All agent actions are messages, observable
   by the e-institution
   An agent should never break the norms
ISLANDER: Performative Structure
Scene conversational graph: Reception Room




          Each arrow is a concrete message
Objectives of the AMELI middleware
    Mediate and facilitate agent communication within
    conversations (scenes)

    Coordinate and enforce:
      To guarantee the correct evolution of each conversation
      (preventing errors made by the participating agents by
      filtering erroneous illocutions, thus protecting the
      institution)
      To guarantee that agents’ movements between scenes
      comply with the specification
      To control which obligations participating agents acquire
      and fulfil
GOVERNORS



                    A1        ...       Ai      ...            An              Agents
                                                                                Layer




                                                                     Public
 Institution
                    G1        ...       Gi      ...           Gn
Specification                                                                  AMELI
   (XML                                                                        Social
                                                                               Layer




                                                                     Private
  format)                                ...
                    IM        SM1   ... S Mm   TM1        ... T Mk
                          -




                                                      -
                               Communication Layer



                INSTITUTION      SCENE          TRANSITION
                 MANAGER        MANAGERS        MANAGERS
AMELI – Agents in Social Layer
    An institution manager that starts the
    institution, authorises agents to enter, and
    controls the creation of scenes
    Scene managers responsible for governing
    scenes (one for scene)
    Transition     managers       control   agents’
    movements between scenes (one for
    transition)
    Governors mediate the interaction of an agent
    with the rest of the agents within the
    institution and control the agents’ obligations
    (one for participating agent)
Organizational Structures
  A pattern of information and control
  relationships between individuals
  Responsible for shaping the types of
  interactions among the agents
  Aids coordination by specifying which
  actions an agent will undertake
  Social structure-based methods impose
  restrictions or norms on the behaviour of
  agents in a certain environment
Sociology and Societies
   Sociology is a discipline that results from an
   evolution of Philosophy in order to describe
   the interactions that arise among the members
   of a group, and the social structures that are
   established
   The aim of any society is to allow its members
   to coexist in a shared environment and pursue
   their respective goals in the presence and/or in
   co-operation with others
   This can also be applied to digital societies
   composed by computational entities (agent
   societies)
Organizational studies (I)

    Organizational studies, organizational
    behaviour, and organizational theory are
    related terms for the academic study of
    organizations
    They have been examined using the
    methods of economics, sociology, political
    science, anthropology and psychology
Organizational studies (II)

 Concepts, abstractions and techniques coming from
 organizational theories and organizational design
 have been used in MAS
   Organization theory is a descriptive discipline, mainly
   focusing on describing and understanding organizational
   functioning
   Organization design is a normative, design-oriented
   discipline that aims to produce the frameworks and tools
   required to create effective organizations
Organization design
   Organization design involves the creation of
   roles, processes and formal reporting
   relationships in an organization
   One can distinguish between two phases in an
   organization design process:
     Strategic grouping, which establishes the overall
     structure of the organization (its main sub-units and
     their relationships), and
     Operational design, which defines the more detailed
     roles and processes
Social Structures
   In open systems, some kind of structure should
   be defined in order to ease coordination in a
   distributed control scenario
   A good option taken from human and animal
   interactions is the definition of social structures
   Social structures define a social level where
   the multi-agent system is seen as a society of
   entities in order to enhance the coordination of
   agent activities (such as message passing
   management and the allocation of tasks and
   resources) by defining structured patterns of
   behaviour
Social Structures - Aim
   Social structures reduce the danger of
   combinatorial explosion in dealing with the
   problems of agent cognition, cooperation and
   control, as they impose restrictions to the agents’
   actions

   These restrictions have a positive effect, as they:
     avoid many potential conflicts, or ease their resolution
     make easier for a given agent to foresee and model
     other agents’ behaviour in a closed environment and fit
     its own behaviour accordingly
Social Strucs. - Organizational classification
    Markets, where agents are self-interested, driven
    completely by their own goals. Interaction in
    markets occurs through communication and
    negotiation

    Networks, where coalitions of self-interested
    agents agree to collaborate in order to achieve a
    mutual goal. Coordination is achieved by mutual
    interest, possibly using trusted third parties

    Hierarchies, where agents are fully cooperative,
    and coordination is achieved through command
    and control lines
Social Structures
Organizational classification




   This classification is useful at the design stage, as
   it tries to motivate the choice of one structure
   based on its appropriateness for a specific
   environment
Market structures
 They are well-suited for
 environments where the
 main purpose is the
 exchange of some goods
 There are agents that
 provide services, agents
 that require services (and
 pay for them), and
 intermediate agents
Network structures

 They are well-suited for
 environments where
 (dynamic) collaboration
 among parties is
 needed
 There are contracts
 established between
 the agents of the
 system
Hierarchies
   Hierarchical structures
   are well-suited for
   environments where the
   society’s purpose is the
   efficient production of
   some kind of results or
   goods. Agents are
   specialised in concrete
   tasks
Social abstractions (I) - Role

    Roles identify activities and services
    necessary to achieve social objectives and
    enable to abstract from the specific individuals
    that will eventually perform them
       From the society design perspective, roles
       provide the building blocks for the agent
       systems that can perform the role
       From the agent design perspective, roles
       specify the expectations of the society with
       respect to the agent’s activity in the society
Social abstractions (II) : Role Dependency

   Role dependency between two roles means
   that one role is dependent on another role for
   the realization of its objectives.
     Societies establish dependencies and power
     relations between roles, indicating relationships
     between roles
     These relationships describe how actors can interact
     and contribute to the realization of the objectives of
     each other. That is, an objective of a role can be
     delegated to, or requested from, other roles
Agent Societies – Characteristics (I)
  Role models reflect social competence of agents
    Modelled by rights and obligations
    Influence agent behaviour
  Role models allow to ensure some global system
  characteristics while also preserving individual
  flexibility
    Explicit rights and obligations allow to commit to specific
    roles
    Roles guarantee global behaviour
    Role descriptions are represented by formal models
Agent Societies – Characteristics (II)
 Interaction models reflect workflows and
 business processes
   Explicit procedures and access requirements
   Scenes descriptions are formally specified, which
   allows verification
Example of organisation structure
    Production of different types of cars
    within a factory
    It involves several kinds of actors:
    engineers, designers, salesmen,
    different types of managers
Coordination Structure 1
Product Hierarchy


      Product Manager I                   Product Manager 2




                                     Designer    Engineer     Salesman
 Designer    Engineer     Salesman
Product hierarchy
 There is a dedicated team for each product (type
 of car) to be produced
 Easy coordination within each product team
 There may be global inefficiencies
   Repetition of design and engineering tasks in different
   products
   A salesman may be specialised in a single product,
   without enough knowledge/abilities to talk to a costumer,
   identify his requirements and suggest the best product for
   him
   There might be a “global manager” trying to provide some
   global communication and coordination
 It might be a good option if products are quite
 different from each other
Coordination Structure 2
Functional Hierarchy

              Product Manager (several products)




  Design                                              Sales
                       Engineering
 Manager                                             Manager
                        Manager




 Designers                                         Salesmen
                       Engineers
Functional hierarchy (I)

  Actors with the same role work together
  under the supervision of a manager
  A general product manager coordinates all
  the activities of all the departments
  Firemen/policemen/ambulances in the
  practical exercise
Functional hierarchy (II)
  The specialised actors can work in tasks
  reusable in different products (e.g. designing
  and engineering the air-conditioning system)
  The resources in each department can be
  easily shared by its members
  Much work concentrated in the global product
  manager, who must supervise the work of the
  whole system
  It can be a good option if the different
  products are very interrelated
Coordination Structures 3
         Product + Functional Hierarchy

        Product Manager 1   Product Manager 2   Product Manager 3




Functional
Managers

              Design          Engineering              Sales
             Manager           Manager                Manager




             Designers        Engineers            Salesmen
Product and functional hierarchy (I)
  There are specialised departments, with a
  manager for each of them (department head,
  or functional manager)
  There is a product manager for each product,
  who talks to the functional managers
  Functional managers act like brokers
  Brokers are in contact with possible ”workers”
  and will choose the best for each task
Product and functional hierarchy (II)
  Few connections and communication messages
  are required
  Quite similar to the functional model
  A lot of work for functional managers
    Receive requests from several product managers
    Coordinate the work of a team of agents
    Identify common subtasks, manage shared resources
  The failure of one product manager does not affect
  the others
Coordination Structure 4
 Flat Structure


Product Manager 1   Product Manager 2   Product Manager 3




   Designers          Engineers          Salesmen
Flat structure
 There is a product manager for each product, who
 talks directly to the low-level workers, without
 intermediate steps
 A product manager may have to communicate with
 many different agents, and these agents have
 different abilities/expertise/vocabulary
 Furthermore, there may be inefficiencies in the
 global behaviour
   A designer could have work in 2 products, while another
   designer does not have any work
   Two engineers could be working in similar problems in two
   different products
   Difficult to solve even with a high-level global coordinator
Organizational Structures - Critique

 Useful when there are master/slave
 relationships in the MAS.
 Control over the slaves actions – mitigates
 against benefits of DAI such as reliability,
 concurrency
 In some cases it presumes that at least
 one agent has global overview – an
 unrealistic assumption in MAS
Summary of Organisations
Focus on a structure / context for
coordination
Consider different types of structures:
  Peer systems, markets, hierarchies, etc.
Are concerned with streamlining or “hard-
wiring” certain patterns which help
coordination in distributed problem solving
Structure of the MAS - exercise
Comments on the practical exercise

 Implicit cooperation
   The functional organisation of the system has
   been chosen by each working group
   This structure limits the coordination
   possibilities, and determines the communication
   flows between the different types of agents
     For instance, an ambulance cannot talk directly
     with a police car, or team coordinators cannot talk
     between them (in principle)
Readings for this week

 Sections 8.6.3/4 of the book An introduction
 to MultiAgent Systems (M. Wooldridge, 2nd
 edition)
 Article: Ant-based load balancing in
 telecommunications networks
 Article: The organ allocation process: a
 natural extension of the Carrel agent-
 mediated electronic institution

Weitere ähnliche Inhalte

Was ist angesagt?

An agent based approach for building complex software systems
An agent based approach for building complex software systemsAn agent based approach for building complex software systems
An agent based approach for building complex software systems
Icaro Santos
 

Was ist angesagt? (15)

Hci unit 4( final)
Hci unit 4( final)Hci unit 4( final)
Hci unit 4( final)
 
An agent based approach for building complex software systems
An agent based approach for building complex software systemsAn agent based approach for building complex software systems
An agent based approach for building complex software systems
 
IFI7159 M0
IFI7159 M0IFI7159 M0
IFI7159 M0
 
Hci md exam
Hci md examHci md exam
Hci md exam
 
The NeuroPsychology of Presence
The NeuroPsychology of PresenceThe NeuroPsychology of Presence
The NeuroPsychology of Presence
 
A participatory modelling method for co-designing a shared representation of ...
A participatory modelling method for co-designing a shared representation of ...A participatory modelling method for co-designing a shared representation of ...
A participatory modelling method for co-designing a shared representation of ...
 
HCI - Chapter 4
HCI - Chapter 4HCI - Chapter 4
HCI - Chapter 4
 
IFI7159 M3
IFI7159 M3IFI7159 M3
IFI7159 M3
 
C0353018026
C0353018026C0353018026
C0353018026
 
Software Agent Oriented Programming
Software Agent Oriented Programming Software Agent Oriented Programming
Software Agent Oriented Programming
 
Beyond believable agents - employing AI for improving game like simulations f...
Beyond believable agents - employing AI for improving game like simulations f...Beyond believable agents - employing AI for improving game like simulations f...
Beyond believable agents - employing AI for improving game like simulations f...
 
Pattern-based competence management
Pattern-based competence managementPattern-based competence management
Pattern-based competence management
 
Hci activity#1
Hci activity#1Hci activity#1
Hci activity#1
 
MuMe Slide M. Wolpers 18 Nov
MuMe Slide M. Wolpers 18 NovMuMe Slide M. Wolpers 18 Nov
MuMe Slide M. Wolpers 18 Nov
 
USER EXPERIENCE AND DIGITALLY TRANSFORMED/CONVERTED EMOTIONS
USER EXPERIENCE AND DIGITALLY TRANSFORMED/CONVERTED EMOTIONSUSER EXPERIENCE AND DIGITALLY TRANSFORMED/CONVERTED EMOTIONS
USER EXPERIENCE AND DIGITALLY TRANSFORMED/CONVERTED EMOTIONS
 

Andere mochten auch

הגדרות פרטיות בפייסבוק
הגדרות פרטיות בפייסבוקהגדרות פרטיות בפייסבוק
הגדרות פרטיות בפייסבוק
012 Smile
 
ניהול סימניות בגוגל כרום
ניהול סימניות בגוגל כרוםניהול סימניות בגוגל כרום
ניהול סימניות בגוגל כרום
012 Smile
 
מצגת 012 netstat
מצגת 012   netstatמצגת 012   netstat
מצגת 012 netstat
012 Smile
 
The West And The World Social Blogger Rafael
The West And The World Social Blogger RafaelThe West And The World Social Blogger Rafael
The West And The World Social Blogger Rafael
Daniel
 

Andere mochten auch (20)

MAS course Lect13 industrial applications
MAS course Lect13 industrial applicationsMAS course Lect13 industrial applications
MAS course Lect13 industrial applications
 
Agent architectures
Agent architecturesAgent architectures
Agent architectures
 
Introduction to agents and multi-agent systems
Introduction to agents and multi-agent systemsIntroduction to agents and multi-agent systems
Introduction to agents and multi-agent systems
 
Lecture 4- Agent types
Lecture 4- Agent typesLecture 4- Agent types
Lecture 4- Agent types
 
MAS course - Lect11 - URV applications
MAS course - Lect11 - URV applicationsMAS course - Lect11 - URV applications
MAS course - Lect11 - URV applications
 
Lect 8-auctions
Lect 8-auctionsLect 8-auctions
Lect 8-auctions
 
MAS course - Lect12 - URV health care applications
MAS course - Lect12 - URV health care applicationsMAS course - Lect12 - URV health care applications
MAS course - Lect12 - URV health care applications
 
MAS course - Lect 9
MAS course - Lect 9 MAS course - Lect 9
MAS course - Lect 9
 
Lect6-An introduction to ontologies and ontology development
Lect6-An introduction to ontologies and ontology developmentLect6-An introduction to ontologies and ontology development
Lect6-An introduction to ontologies and ontology development
 
ECAI 2014 poster - Unsupervised semantic clustering of Twitter hashtags
ECAI 2014 poster - Unsupervised semantic clustering of Twitter hashtagsECAI 2014 poster - Unsupervised semantic clustering of Twitter hashtags
ECAI 2014 poster - Unsupervised semantic clustering of Twitter hashtags
 
הגדרות פרטיות בפייסבוק
הגדרות פרטיות בפייסבוקהגדרות פרטיות בפייסבוק
הגדרות פרטיות בפייסבוק
 
model's
model'smodel's
model's
 
ניהול סימניות בגוגל כרום
ניהול סימניות בגוגל כרוםניהול סימניות בגוגל כרום
ניהול סימניות בגוגל כרום
 
Drug Prescriptions
Drug PrescriptionsDrug Prescriptions
Drug Prescriptions
 
מצגת 012 netstat
מצגת 012   netstatמצגת 012   netstat
מצגת 012 netstat
 
Otchet
OtchetOtchet
Otchet
 
TNM классификация немелкоклеточного рака легкого
TNM классификация немелкоклеточного рака легкогоTNM классификация немелкоклеточного рака легкого
TNM классификация немелкоклеточного рака легкого
 
TNM классификация рака легкого (6 редакция)
TNM классификация рака легкого (6 редакция)TNM классификация рака легкого (6 редакция)
TNM классификация рака легкого (6 редакция)
 
The West And The World Social Blogger Rafael
The West And The World Social Blogger RafaelThe West And The World Social Blogger Rafael
The West And The World Social Blogger Rafael
 
Iceberg, Dead Ahead - Lessons From Aviation Disasters
Iceberg, Dead Ahead - Lessons From Aviation DisastersIceberg, Dead Ahead - Lessons From Aviation Disasters
Iceberg, Dead Ahead - Lessons From Aviation Disasters
 

Ähnlich wie MAS Course - Lect10 - coordination

Rzevsky agent models of large systems
Rzevsky  agent models of large systemsRzevsky  agent models of large systems
Rzevsky agent models of large systems
Masha Rudnichenko
 
Multiagent systems (and their use in industry)
Multiagent systems (and their use in industry)Multiagent systems (and their use in industry)
Multiagent systems (and their use in industry)
Marc-Philippe Huget
 
Semantic Interoperation of Information Systems by Evolving Ontologies through...
Semantic Interoperation of Information Systems by Evolving Ontologies through...Semantic Interoperation of Information Systems by Evolving Ontologies through...
Semantic Interoperation of Information Systems by Evolving Ontologies through...
Christophe Debruyne
 

Ähnlich wie MAS Course - Lect10 - coordination (20)

c27_mas.ppt
c27_mas.pptc27_mas.ppt
c27_mas.ppt
 
Academic Course: 03 Autonomic Multi-Agent Systems
Academic Course: 03 Autonomic Multi-Agent SystemsAcademic Course: 03 Autonomic Multi-Agent Systems
Academic Course: 03 Autonomic Multi-Agent Systems
 
Presentation_DAI
Presentation_DAIPresentation_DAI
Presentation_DAI
 
Network and spatial analysis for forest governance
Network and spatial analysis  for forest governanceNetwork and spatial analysis  for forest governance
Network and spatial analysis for forest governance
 
Contribution Of Sociotechnical Systems Theory Concepts To A Framework Of Terr...
Contribution Of Sociotechnical Systems Theory Concepts To A Framework Of Terr...Contribution Of Sociotechnical Systems Theory Concepts To A Framework Of Terr...
Contribution Of Sociotechnical Systems Theory Concepts To A Framework Of Terr...
 
Contribution of Socio-Technical Systems Theory Concepts to a Framework of Ter...
Contribution of Socio-Technical Systems Theory Concepts to a Framework of Ter...Contribution of Socio-Technical Systems Theory Concepts to a Framework of Ter...
Contribution of Socio-Technical Systems Theory Concepts to a Framework of Ter...
 
Agent Interaction Agents
Agent Interaction AgentsAgent Interaction Agents
Agent Interaction Agents
 
Rzevsky agent models of large systems
Rzevsky  agent models of large systemsRzevsky  agent models of large systems
Rzevsky agent models of large systems
 
Mathmodsocsys3 090827155611-phpapp02
Mathmodsocsys3 090827155611-phpapp02Mathmodsocsys3 090827155611-phpapp02
Mathmodsocsys3 090827155611-phpapp02
 
Coordination of Complex Socio-technical Systems: Challenges and Opportunities
Coordination of Complex Socio-technical Systems: Challenges and OpportunitiesCoordination of Complex Socio-technical Systems: Challenges and Opportunities
Coordination of Complex Socio-technical Systems: Challenges and Opportunities
 
Agent-Based Modeling for Sociologists
Agent-Based Modeling for SociologistsAgent-Based Modeling for Sociologists
Agent-Based Modeling for Sociologists
 
Systems Thinking an Introduction
Systems Thinking an IntroductionSystems Thinking an Introduction
Systems Thinking an Introduction
 
Swarm intel
Swarm intelSwarm intel
Swarm intel
 
Multiagent systems (and their use in industry)
Multiagent systems (and their use in industry)Multiagent systems (and their use in industry)
Multiagent systems (and their use in industry)
 
Sna based reasoning for multiagent
Sna based reasoning for multiagentSna based reasoning for multiagent
Sna based reasoning for multiagent
 
Community detection
Community detectionCommunity detection
Community detection
 
Semantic Interoperation of Information Systems by Evolving Ontologies through...
Semantic Interoperation of Information Systems by Evolving Ontologies through...Semantic Interoperation of Information Systems by Evolving Ontologies through...
Semantic Interoperation of Information Systems by Evolving Ontologies through...
 
Networks and Organisational Work
Networks and Organisational WorkNetworks and Organisational Work
Networks and Organisational Work
 
Improving Knowledge Handling by building intellegent social systems
Improving Knowledge Handling by building intellegent social systemsImproving Knowledge Handling by building intellegent social systems
Improving Knowledge Handling by building intellegent social systems
 
Research Trends in Nature-inspired Coordination Models
Research Trends in Nature-inspired Coordination ModelsResearch Trends in Nature-inspired Coordination Models
Research Trends in Nature-inspired Coordination Models
 

Mehr von Antonio Moreno

On the application of multi-agent systems in Health Care
On the application of multi-agent systems in Health CareOn the application of multi-agent systems in Health Care
On the application of multi-agent systems in Health Care
Antonio Moreno
 

Mehr von Antonio Moreno (8)

Dynamic learning of keyword-based preferences for news recommendation (WI-2014)
Dynamic learning of keyword-based preferences for news recommendation (WI-2014)Dynamic learning of keyword-based preferences for news recommendation (WI-2014)
Dynamic learning of keyword-based preferences for news recommendation (WI-2014)
 
Automatic and unsupervised topic discovery in social networks
Automatic and unsupervised topic discovery in social networksAutomatic and unsupervised topic discovery in social networks
Automatic and unsupervised topic discovery in social networks
 
Artificial Intelligence Master URV-UPC-UB
Artificial Intelligence Master URV-UPC-UBArtificial Intelligence Master URV-UPC-UB
Artificial Intelligence Master URV-UPC-UB
 
URV Master on Computer Engineering: Computer Security and Inteligent Systems
URV Master on Computer Engineering: Computer Security and Inteligent SystemsURV Master on Computer Engineering: Computer Security and Inteligent Systems
URV Master on Computer Engineering: Computer Security and Inteligent Systems
 
EnoSigTur: recomanació personalitzada d'activitats enoturístiques
EnoSigTur: recomanació personalitzada d'activitats enoturístiquesEnoSigTur: recomanació personalitzada d'activitats enoturístiques
EnoSigTur: recomanació personalitzada d'activitats enoturístiques
 
On the application of multi-agent systems in Health Care
On the application of multi-agent systems in Health CareOn the application of multi-agent systems in Health Care
On the application of multi-agent systems in Health Care
 
Multi-agent systems applied in Health Care
Multi-agent systems applied in Health CareMulti-agent systems applied in Health Care
Multi-agent systems applied in Health Care
 
Artificial Intelligence techniques in Tourism at URV
Artificial Intelligence techniques in Tourism at URVArtificial Intelligence techniques in Tourism at URV
Artificial Intelligence techniques in Tourism at URV
 

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost SavingRepurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
GenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdfGenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdf
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organization
 
Top 10 Most Downloaded Games on Play Store in 2024
Top 10 Most Downloaded Games on Play Store in 2024Top 10 Most Downloaded Games on Play Store in 2024
Top 10 Most Downloaded Games on Play Store in 2024
 
AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of Terraform
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CV
 
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 

MAS Course - Lect10 - coordination

  • 1. LECTURE 10: Cooperation in MAS (IV): implicit methods Artificial Intelligence II – Multi-Agent Systems Introduction to Multi-Agent Systems URV, Winter-Spring 2010
  • 2. Outline of the lecture Implicit cooperation in MAS Indirect cooperation through the environment Societal views of MAS Electronic institutions Organizational structures
  • 3. Coordination [recall past lectures] An activity is a set of potential operations an actor (an agent playing a certain role) can perform, with the aim of achieving a given goal or set of goals Coordination could be defined as the process of managing dependencies between activities. By such process an agent reasons about its local actions and the foreseen actions that other agents may perform, with the aim to make the community to behave globally in a coherent manner
  • 4. Cooperation hierarchy [last lectures] MAS Independent Cooperative Self-interested Benevolent Discrete Emergent With Without communication - communication - Explicit Implicit Reactive systems Deliberative Negotiators Partial Global Auctions Planning Coalition Voting formation Contract Net
  • 5. Implicit cooperation A group of distributed cooperative agents behaves in a socially coordinated way in the resolution of a global problem without an explicit exchange of communication messages In many cases the environment acts as the (indirect) interaction mechanism
  • 6. Motivation (I) Cases in which explicit coordination cannot be applied: Speed: it takes too long to communicate with others – by then the opportunities are missed E.g. Football game – simple signals may work, but lengthy explanations don't... In general, very dynamic environments Security: not wanting others to know what your plans are
  • 7. Motivation (II) Complexity: some agents may be too simple to deal with the complexity of generating and understanding complex plans Reactive rule-based robots Complexity of Partial Global Planning or coalition formation Lack of a communication channel: there may actually be no way to communicate Physical robots with limited communication range
  • 8. Options for implicit cooperation Observe the behaviour of the other agents, and react accordingly Indirect cooperation through the effects on the environment of the actions of each agent Imposing a structure on the MAS
  • 9. Emergent Coordination [recall past lectures] Coordination in cases where: There is no communication between agents There is no mechanism for enforcing a-priori social rules / laws Agents have their own agenda/goals The resulting coordination is emergent and cannot be said to be based on joint plans or intentions
  • 10. Basic difference Emergent coordination: agents are self- interested, they do not care about the other agents in the system, there isn’t any high level design of the emergent behaviour Implicit coordination (also giving rise to emergent coordinated global behaviour): although agents do not communicate with each other, the designer of the system intends to provoke the emergence of the socially intelligent problem solving activities
  • 11. Implicit coordination example: Network Routing Network Routing problems are challenging. Solutions need to be: Dynamic Robust Network of N nodes, L links. Traffic flows as packets traverse the network There are protocols that compute cumulative shortest path measures
  • 13. Network Ants Ants randomly explore the network until they find a specific node They mark the traversed paths with “pheromone” Ants seeking destinations follow pheromone trails Pheromones degrade over time Robust Stable Gradual Change
  • 14. Pheromone tables Each node contains a table of probabilities (pheromone table) for each possible destination in the network In a 30-nodes network, each node keeps 29 tables The entries on the tables are the probabilities which influence the ants’ selection of the next node on the way to their destination node Pheromone laying = updating probabilities
  • 15. Pheromone tables example A network with 6 nodes, node 1 is connected with nodes 2, 4 and 5. Next node The pheromone tables in 2 4 5 node 1 would look like this: For instance, if an ant 2 0.90 0.05 0.05 arrives at node 1 and wants to go to node 3, the most 3 0.25 0.60 0.15 probable route is through Destination 4 0.10 0.85 0.05 node 4 (but it may also node 5 0.10 0.10 0.80 decide to go through nodes 2 or 5) 6 0.40 0.30 0.30
  • 16. Simulation (I) At each step, ants can be launched from any node in the network, with a random destination node Ants move from node to node, selecting the next node to move to according to the probabilities in the pheromone tables for their destination node Pheromone tables are initialized with random values
  • 17. Simulation (II) When ants arrive at a node, they update the probabilities of that node’s pheromone table entries corresponding to their source node They alter the table to increase the probability pointing to their previous node Ants moving away from their source node can only directly affect those ants for which it is the destination node
  • 18. Pheromone laying example An ant has to go from node 3 to node 2; in the way, it travels from node 4 to node 1 First, it modifies the table in node 1 corresponding to node 3, increasing the probability of selecting the link to node 4 After that, it selects the next node randomly according to the probabilities of the table in node 1 corresponding to node 2 3 … 4 1 … 2
  • 19. Increasing/decreasing pheromones Pheromones are increased with the following formula p = (p_old + Δ(p)) / (1 + Δ(p)) As all the entries must add up to 1, the other entries have to be decreased as follows p = p_old / (1 + Δ(p)) Note that probabilities may never be 0
  • 20. Basic ideas of implicit cooperation Agents do not talk to each other directly Agents can modify the environment, and these modifications influence the behaviour of the other agents in the system All the agents contribute towards a useful global behaviour of the community
  • 21. Reasoning mechanisms for coordination Thinking about individual agents Methods that allow building a model of the other agents of the system Thinking about the whole agent’s society Methods that try to impose some kind of rules/laws/structure/organisation in the multi- agent system
  • 22. Agent Modelling (I) Even if you cannot talk to the other agents you may still want to reason about them Main methods: Recursive Modelling Methods Assume the others have a similar structure to you – and may have a model of you... Try to deduce their beliefs/desires/intentions from their actions on the environment
  • 23. Agent modelling (II) Plan Recognition Analyse the sequences of activities of other agents and try to discover their plans (and, from them, identify the potential end goals of their actual actions) Game Playing / Game Tree Search: Modelling opponents For example, using minimax search [Recall Game Theory in Artificial Intelligence]
  • 24. Thinking about Society Common approaches include: Social Laws: global rules which agents follow and lead to “coherent behaviour”, either instilled in the agent or communicated when entering the environment (e.g. - “driving on the right hand side”) Social Power Relations: a theory of dependence relations, in particular to model goal adoption (e.g. carrying out work on behalf of a superior) Electronic Institutions Organizational structures
  • 25. Institutions as Social Structures Social Structures define a social level to enhance coordination by means of interaction patterns Institutions are a kind of social structure where a corpora of constraints shape the behaviour of the members of a group
  • 26. Institution components The definition of a (human) Institution usually includes: Norms about the interactions Conventions: acceptable (and unacceptable) actions within the institution Procedures and protocols to be followed
  • 27. e-Institutions An e-Institution is the computational model of an institution through The specification of the institution’s norms in some suitable formalism The formal specification of the institution’s admissible procedures and protocols, which follow the established conventions
  • 28. E-Institutions and MAS In the context of MAS, e-institutions: reduce uncertainty of other agents’ behaviour reduce misunderstanding in interaction allow agents to foresee the outcome of an interaction simplify the decision making process (by reducing the possible actions) Agent behaviour guided by Norms
  • 29. Why a Language for Norms? Laws, Laws, [Natural Language] regulations regulations too abstract and vague Language for norms Language for norms [Formal Language] more concrete (Formal & Computational) (Formal & Computational) Electronic Institutions Normative Agents Norms in Norm enforcement deliberation cycle mechanisms
  • 30. Influence of norms in the BDI deliberation cycle input Agent sensors perception E state How is the N world now? V I What if I perform R O KB action A? N M Which action do E I choose? N T goals actuators norms (obligations, permissions...) action
  • 31. AMELI (I) AMELI is an institution middleware that is based in a formal electronic institution specification tool (ISLANDER), developed at IIIA The ISLANDER framework is composed of: A Dialogical Framework Linguistic and social structure (roles) to give meaning to agent interactions, communication language A Performative Structure scenes and relationships between scenes (e.g. precedence) Rules Conventions to be followed, social commitments
  • 32. AMELI (II) Two hypotheses: All agent actions are messages, observable by the e-institution An agent should never break the norms
  • 34. Scene conversational graph: Reception Room Each arrow is a concrete message
  • 35.
  • 36. Objectives of the AMELI middleware Mediate and facilitate agent communication within conversations (scenes) Coordinate and enforce: To guarantee the correct evolution of each conversation (preventing errors made by the participating agents by filtering erroneous illocutions, thus protecting the institution) To guarantee that agents’ movements between scenes comply with the specification To control which obligations participating agents acquire and fulfil
  • 37. GOVERNORS A1 ... Ai ... An Agents Layer Public Institution G1 ... Gi ... Gn Specification AMELI (XML Social Layer Private format) ... IM SM1 ... S Mm TM1 ... T Mk - - Communication Layer INSTITUTION SCENE TRANSITION MANAGER MANAGERS MANAGERS
  • 38. AMELI – Agents in Social Layer An institution manager that starts the institution, authorises agents to enter, and controls the creation of scenes Scene managers responsible for governing scenes (one for scene) Transition managers control agents’ movements between scenes (one for transition) Governors mediate the interaction of an agent with the rest of the agents within the institution and control the agents’ obligations (one for participating agent)
  • 39. Organizational Structures A pattern of information and control relationships between individuals Responsible for shaping the types of interactions among the agents Aids coordination by specifying which actions an agent will undertake Social structure-based methods impose restrictions or norms on the behaviour of agents in a certain environment
  • 40. Sociology and Societies Sociology is a discipline that results from an evolution of Philosophy in order to describe the interactions that arise among the members of a group, and the social structures that are established The aim of any society is to allow its members to coexist in a shared environment and pursue their respective goals in the presence and/or in co-operation with others This can also be applied to digital societies composed by computational entities (agent societies)
  • 41. Organizational studies (I) Organizational studies, organizational behaviour, and organizational theory are related terms for the academic study of organizations They have been examined using the methods of economics, sociology, political science, anthropology and psychology
  • 42. Organizational studies (II) Concepts, abstractions and techniques coming from organizational theories and organizational design have been used in MAS Organization theory is a descriptive discipline, mainly focusing on describing and understanding organizational functioning Organization design is a normative, design-oriented discipline that aims to produce the frameworks and tools required to create effective organizations
  • 43. Organization design Organization design involves the creation of roles, processes and formal reporting relationships in an organization One can distinguish between two phases in an organization design process: Strategic grouping, which establishes the overall structure of the organization (its main sub-units and their relationships), and Operational design, which defines the more detailed roles and processes
  • 44. Social Structures In open systems, some kind of structure should be defined in order to ease coordination in a distributed control scenario A good option taken from human and animal interactions is the definition of social structures Social structures define a social level where the multi-agent system is seen as a society of entities in order to enhance the coordination of agent activities (such as message passing management and the allocation of tasks and resources) by defining structured patterns of behaviour
  • 45. Social Structures - Aim Social structures reduce the danger of combinatorial explosion in dealing with the problems of agent cognition, cooperation and control, as they impose restrictions to the agents’ actions These restrictions have a positive effect, as they: avoid many potential conflicts, or ease their resolution make easier for a given agent to foresee and model other agents’ behaviour in a closed environment and fit its own behaviour accordingly
  • 46. Social Strucs. - Organizational classification Markets, where agents are self-interested, driven completely by their own goals. Interaction in markets occurs through communication and negotiation Networks, where coalitions of self-interested agents agree to collaborate in order to achieve a mutual goal. Coordination is achieved by mutual interest, possibly using trusted third parties Hierarchies, where agents are fully cooperative, and coordination is achieved through command and control lines
  • 47. Social Structures Organizational classification This classification is useful at the design stage, as it tries to motivate the choice of one structure based on its appropriateness for a specific environment
  • 48. Market structures They are well-suited for environments where the main purpose is the exchange of some goods There are agents that provide services, agents that require services (and pay for them), and intermediate agents
  • 49. Network structures They are well-suited for environments where (dynamic) collaboration among parties is needed There are contracts established between the agents of the system
  • 50. Hierarchies Hierarchical structures are well-suited for environments where the society’s purpose is the efficient production of some kind of results or goods. Agents are specialised in concrete tasks
  • 51. Social abstractions (I) - Role Roles identify activities and services necessary to achieve social objectives and enable to abstract from the specific individuals that will eventually perform them From the society design perspective, roles provide the building blocks for the agent systems that can perform the role From the agent design perspective, roles specify the expectations of the society with respect to the agent’s activity in the society
  • 52. Social abstractions (II) : Role Dependency Role dependency between two roles means that one role is dependent on another role for the realization of its objectives. Societies establish dependencies and power relations between roles, indicating relationships between roles These relationships describe how actors can interact and contribute to the realization of the objectives of each other. That is, an objective of a role can be delegated to, or requested from, other roles
  • 53. Agent Societies – Characteristics (I) Role models reflect social competence of agents Modelled by rights and obligations Influence agent behaviour Role models allow to ensure some global system characteristics while also preserving individual flexibility Explicit rights and obligations allow to commit to specific roles Roles guarantee global behaviour Role descriptions are represented by formal models
  • 54. Agent Societies – Characteristics (II) Interaction models reflect workflows and business processes Explicit procedures and access requirements Scenes descriptions are formally specified, which allows verification
  • 55. Example of organisation structure Production of different types of cars within a factory It involves several kinds of actors: engineers, designers, salesmen, different types of managers
  • 56. Coordination Structure 1 Product Hierarchy Product Manager I Product Manager 2 Designer Engineer Salesman Designer Engineer Salesman
  • 57. Product hierarchy There is a dedicated team for each product (type of car) to be produced Easy coordination within each product team There may be global inefficiencies Repetition of design and engineering tasks in different products A salesman may be specialised in a single product, without enough knowledge/abilities to talk to a costumer, identify his requirements and suggest the best product for him There might be a “global manager” trying to provide some global communication and coordination It might be a good option if products are quite different from each other
  • 58. Coordination Structure 2 Functional Hierarchy Product Manager (several products) Design Sales Engineering Manager Manager Manager Designers Salesmen Engineers
  • 59. Functional hierarchy (I) Actors with the same role work together under the supervision of a manager A general product manager coordinates all the activities of all the departments Firemen/policemen/ambulances in the practical exercise
  • 60. Functional hierarchy (II) The specialised actors can work in tasks reusable in different products (e.g. designing and engineering the air-conditioning system) The resources in each department can be easily shared by its members Much work concentrated in the global product manager, who must supervise the work of the whole system It can be a good option if the different products are very interrelated
  • 61. Coordination Structures 3 Product + Functional Hierarchy Product Manager 1 Product Manager 2 Product Manager 3 Functional Managers Design Engineering Sales Manager Manager Manager Designers Engineers Salesmen
  • 62. Product and functional hierarchy (I) There are specialised departments, with a manager for each of them (department head, or functional manager) There is a product manager for each product, who talks to the functional managers Functional managers act like brokers Brokers are in contact with possible ”workers” and will choose the best for each task
  • 63. Product and functional hierarchy (II) Few connections and communication messages are required Quite similar to the functional model A lot of work for functional managers Receive requests from several product managers Coordinate the work of a team of agents Identify common subtasks, manage shared resources The failure of one product manager does not affect the others
  • 64. Coordination Structure 4 Flat Structure Product Manager 1 Product Manager 2 Product Manager 3 Designers Engineers Salesmen
  • 65. Flat structure There is a product manager for each product, who talks directly to the low-level workers, without intermediate steps A product manager may have to communicate with many different agents, and these agents have different abilities/expertise/vocabulary Furthermore, there may be inefficiencies in the global behaviour A designer could have work in 2 products, while another designer does not have any work Two engineers could be working in similar problems in two different products Difficult to solve even with a high-level global coordinator
  • 66. Organizational Structures - Critique Useful when there are master/slave relationships in the MAS. Control over the slaves actions – mitigates against benefits of DAI such as reliability, concurrency In some cases it presumes that at least one agent has global overview – an unrealistic assumption in MAS
  • 67. Summary of Organisations Focus on a structure / context for coordination Consider different types of structures: Peer systems, markets, hierarchies, etc. Are concerned with streamlining or “hard- wiring” certain patterns which help coordination in distributed problem solving
  • 68. Structure of the MAS - exercise
  • 69. Comments on the practical exercise Implicit cooperation The functional organisation of the system has been chosen by each working group This structure limits the coordination possibilities, and determines the communication flows between the different types of agents For instance, an ambulance cannot talk directly with a police car, or team coordinators cannot talk between them (in principle)
  • 70. Readings for this week Sections 8.6.3/4 of the book An introduction to MultiAgent Systems (M. Wooldridge, 2nd edition) Article: Ant-based load balancing in telecommunications networks Article: The organ allocation process: a natural extension of the Carrel agent- mediated electronic institution