=Paper=
{{Paper
|id=None
|storemode=property
|title=Programming Open Systems with Agents, Environments and Organizations
|pdfUrl=https://ceur-ws.org/Vol-621/paper06.pdf
|volume=Vol-621
|dblpUrl=https://dblp.org/rec/conf/woa/PiuntiRBH10
}}
==Programming Open Systems with Agents, Environments and Organizations==
https://ceur-ws.org/Vol-621/paper06.pdf
Programming Open Systems with
Agents, Environments and Organizations
Michele Piunti, Olivier Boissier Jomi F. Hübner
Alessandro Ricci Ecole Nationale Supérieure des Mines Universidade Regional de Blumenau
Università di Bologna St-Etienne, France Blumenau, SC - Brazil
Sede di Cesena boissier@emse.fr jomi@inf.furb.br
{michele.piunti,a.ricci}@unibo.it
Abstract—MAS research pushes the notion of openness re- Such a multifaceted perspective risks to turn systems into a
lated to systems combining heterogeneous computational enti- scattered aggregation of heterogenous elements, while their
ties. Typically, those entities answer to different purposes and interplay, as well as their interaction, is reduced to a problem
functions and their integration is a crucial issue. Starting from
a comprehensive approach in developing agents, organizations of technological interoperability. To prevent this, besides the
and environments, this paper devises an integrated approach and different mechanisms and abstractions that must be considered,
describes a unifying programming model. It introduces the notion there is a strong need of binding these elements together in a
of embodied organization, which is described first focusing on the flexible and clear way.
main entities as separate concerns; and, second, establishing dif- Providing a seamless integration of the above aspects places
ferent interaction styles aimed to seamlessly integrate the various
entities in a coherent system. An integration framework, built on the challenge to conceive the proper integration pattern be-
top of Jason, CArtAgO and Moise (as programming platforms tween several entities and constructs. A main concern is agent
for agents, environments and organizations resp.) is described as awareness, namely the need for agents to exhibit special
a suitable technology to build embodied organizations in practice. abilities and knowledge in order to bring about organizational
and environmental notions—which typically are not native
I. I NTRODUCTION constructs of their architectures [21], [15]. Once the environ-
Agent based approaches consider agents as autonomous en- ment dimension is introduced as an additional dimension, a
tities encapsulating their control, characterized (and specified) second concern is how to connect in a meaningful way the
by epistemic states (beliefs) and motivational states (goals) organizational entities and the environmental ones, thereby
which result in a goal oriented behavior. Recently, organiza- (i) how the organization can ground normative measures as
tion oriented computing in Multi Agent Systems (MAS) has regimentation and obligations in environments, and (ii) how
been advocated as a suitable computation model coping with certain events occurring in environments may affect the global
the complex requirements of socio-technical applications. As organizational configuration. These aspects enlighten a series
indicated by many authors [8], [2], [6], organizations are a of drawbacks on existing approaches, either on the conceptual
powerful tool to build complex systems where computational model and on the programming constructs to be adopted to
agents can autonomously pursue their activities exhibiting build systems in practice.
social attitudes. The organizational dimension is conceived Taking a programming perspective, this work describes an
in terms of functionalities to be exploited by agents, while infrastructural support allowing to seamlessly integrate various
it is assumed to control social activities by monitoring and aspects characterizing an open MAS. In doing so, the notion
changing those functionalities at runtime. Being conceived in of Embodied Organization is detailed, aimed at introducing
terms of human organizations, i.e., being structured in terms each element in the MAS as an integral part of a structured
of norms, roles and global objectives, this perspective assumes infrastructure. In order to reconcile organizations, agents and
an organizational layer aimed at promoting desired coordina- environments, Embodied organization allows developers to
tion, improving control and equilibrium of social dynamics. focus on the main entities as separate concerns, and then
Besides, the need for openness and interoperability requires to establish different interaction styles aimed to seamlessly
to cope with computational environments populated by several integrate the various entities in a coherent system. In particular,
entities, not modellable as agents or organizations, which are the proposed approach defines a series of basic mechanisms
supposed to be concurrently exploited by providing function- related to the interaction model:
alities supporting agents objectives. These aspects are even • How the agents could profitably interact with both or-
more recognized in current ICT, characterized by a massive ganizational and other environmental entities in order to
interplay of self-interested entities (humans therein) developed attain their design objectives;
according to different models, technologies and programming • How the organizational entities could control agent ac-
styles. Not surprisingly, recent approaches introduced environ- tivities and regiment environmental resources in order to
ment as pivotal dimension in MAS development [22], [14]. promote desired equilibrium;
• How environmental changes could affect both organiza- “normative artifact”) is conceived as a centralized environment
tional dynamics and agents activities; that is explicitly conceived as a container of institutional facts,
The rest of the paper is organized as follows: Section II i.e., facts related to the normative/institutional states, and brute
provides a survey of situated organization as proposed by ex- facts, i.e. related to the concrete/ “physical” workplace where
isting works. Starting from the description of the basic entities agents work. To shift facts from the brute dimension to the
characterizing an integrated perspective, Section III presents normative one the system is assumed to handle constitutive
a unified programming model including agents, organizations rules defined on the basis of “count-as” and “sanctioning”
and environments. The notion of Embodied Organization is constructs, which allows the infrastructure to recast brute
detailed in Section IV, while Section V discusses a concrete facts to institutional ones. The mechanism regulating the
programming model to implement it in practice. Finally, Sec- application of “count-as” and “sanctioning” rules is then
tion VI concludes the paper discussing the proposed approach based on a monitoring process which is established as an
and future directions. infrastructural functionality embedded inside the normative
system. Thanks to this mechanism, agents behavior can be
II. O RGANIZATIONS SITUATED IN MAS E NVIRONMENTS automatically regulated through enforcing mechanisms, i.e.
Although early approaches in organization programming without the intervention of organizational agents.
have not been addressed at modeling environments explic- A similar approach is proposed in the work by Tinnemeier
itly, recent trends are investigating the challenge to situate et al. [20], where a normative programming language based on
organizations in concrete computational environments. In what conditional obligations and prohibitions is proposed. Thanks
follows, a survey on related works is discussed, enlightening to the inclusion of the environment dimension in the normative
strengths and drawbacks of existing proposals. system, this work explicitly grounds norms either on institu-
tional states either on specific environmental states. In this case
A. Current Approaches indeed the normative system is also in charge of monitoring
Several agent based approaches allow to implement situ- the outcomes of agent activities as performed in the work
ated organizations instrumenting computational environments environment, in so doing providing a twofold support to the
where social interactions are of concern. A remarkable exam- organizational dimension and to the environmental one.
ple of situated organization is due to Okuyama et al. [12], who With the aim to reconcile physical reality with institutional
proposed the use of “normative objects” as reactive entities dimensions, an integral approach has been proposed with the
inspectable by agents working in “normative places”. Nor- MASQ approach, which introduces a meta-model promoting
mative objects can be exploited by the organization to make an analysis and design of a global systems along several
available information about norms that regulate the behavior conceptual dimensions [19]. The MASQ approach relies on
of agents within the place where such objects can be perceived the less recent AGR model, extended with an explicit support
by agents. Indeed, they are supposed to indicate obligations, to environment as envisaged by the AGRE and AGREEN
prohibitions, rights and are readable pieces of information that [1]. Four dimensions are introduced, ranging from endogenous
agents can get and exploit in computational environments. aspects (related to agent’s mental attitudes) to exogenous
The approach envisages a distributed normative infrastructure aspects (related to environments, society and cultures where
which is assumed to control emergent dynamics and to allow agents are immersed). In this case, the same infrastructure
agents to implicitly interact with a normative institution. The used to deploy organizational entities is also regulated by
mechanism is based on the intuition that the reification of precise rules for interactions between agents and environment
a particular state in a normative place may constitute the entities. The resulting interaction model relies on the theory
realization of a particular institutional fact (e.g., “being on of influences and reactions [9], in the context of which several
a car driver seat makes an agent to play the role driver”). This interaction styles can be established among the heterogenous
basic idea is borrowed from John Searle’s work on speech acts entities dwelling the system.
and social reality [16], [17] Searle envisaged an institutional Besides conceptual and formal integration, few approaches
dimension rising out of collective agreements through special have accounted a programming approach for situated organi-
kind of rules, that he refers as constitutive rules. Those rules zations. By relating situated activities in the workplace, the
constitute (and also regulate) an activity the existence of which Brahms platform endows human work practices and allows
is logically dependent on the rules themselves, thus forming a to represent the relations of people, locations, agent systems,
kind of tautology for what a constitutive rule also defines the communication and information content [18]. Based on exist-
notion that it regulates. In this view, “being on a car driver ing theories of situated action, activity theory and distributed
seat makes an agent to play the role driver” strongly situate cognition, the Brahms language promotes the interplay of
the institutional dimension on the environmental one, both intelligent software agents with humans their organizations.
regulating the concept of role adoption and, at the same time, A similar idea is provided by Situated Electronic Institutions
defining it. (SEI) [4], recently proposed as an extension of Electronic
Constitutive rules in the form X counts as Y in C are Institutions (EI) [7]. Besides providing a runtime management
also at the basis of the formal work proposed by Dastani et of the normative specification of dialogic interactions between
al. [5]. Here a normative infrastructure (which is referred as agents, the notion of observability of environment states is
at the basis of SEI. They are aimed at interceding between could fit all together.
real environments and EI. In this case, special governors, Typically interactions are based on a sub-agentive level, and
namely modelers, allow to bridge environmental structures to are founded on protocols and mechanisms, instead on being
the institution by instrumenting environments with “embod- based on the effective capabilities and functionalities exhib-
ied” devices controlled by the institutional apparatus. Partic- ited by the entities involved in the whole system. Different
ipating agents can, in this case, perform individual actions approaches are provided for the interaction model between
and interactions (either non message based) while operating environment, agents and their organizations. Besides, there
upon concrete devices inside the environment. Besides, SEI is not a clear vision on how environment and organizational
introduces the notion of staff agents, namely organization entities should support agents in their native capabilities, as
aware agents which role is to monitor ongoing activities for instance the ones related to action and perception.
performed by agents which are not under the direct control
of the institution. Staff agents are then assumed to bridge The computational treatments of goals clashes different
the gap between participating agents and the institutional approaches once they are referred to agents and their subjective
dimensions: they typically react to norm violations, possibly goals, and when they are related to organizations and their
ascribing sanctioning and enforcements to disobeying agents. global goals. For instance, approaches as MASQ, ORA4MAS
Institutional control is also introduced by the mean of feedback describe in a rather abstract terms (i) how the subjective
mechanisms aimed at comparing observed properties with and global goals should be fulfilled in practice; (ii) which
certain expected values. On the basis of possible not standard brute state has to be reached in order to consider a goal as
properties detected, an autonomic mechanism specifies how achieved. By considering environments explicitly, either agents
reconfigure the institution in order to re-establish equilibrium. and organizations should be able to ground goals to actual
The ORA4MAS approach [11] proposed a programming environment configurations, thus recognizing the fulfillment of
model for concretely building systems integrating organiza- their objectives once the pursued goals have been reached in
tional functionalities in instrumented work environment. In practice (this approach is adopted, for instance, in [5]). Other
ORA4MAS organizational entities are viewed as artifact based approaches, as for instance ORA4MAS [11], do not assume
infrastructures. Specialized organizational artifacts (OAs) are organizations able to automatically detect the fulfillment of
assumed to encapsulate organizational functions, which can global goals in terms of environment configurations.
be exploited by agents to fulfill their organizational purposes. As for goals, a weak support is provided for grounding
Using artifacts as basic building blocks of organizations, norms in concrete application domains, thus allowing to es-
allows agents to natively interact with the organizational entity tablish how and when a norm has been fulfilled or violated.
at a proper abstraction level, namely without being constrained Furthermore few approachess manage norm lifecycle with
to shape external actions as mechanism-level primitives needed respect to distributed and (highly) dynamic environments. No
to work with middleware objects. The consequence is that the agreement is then established on which kind of monitoring and
infrastructure does not rely on a sort of hidden components, sanctioning mechanisms must be adopted. Some approaches
but the organizational layer is placed beside the agents as a envisage the role of organizational/staff agents [4], other
suitable set of services and functionalities to be dynamically approaches propose the sole automatic regulation provided by
exploited (and created) as an integral part of the MAS work a programmable infrastructure [5], [20].
environment. On the other side, ORA4MAS does not provide Different solutions are provided for defining agent capabil-
an explicit support to environmental resources which are ities, namely which grade of awareness is required for agents
not included in the organizational specification. Two types to exploit the functionalities provided by the organizational
of agents are assumed to evolve in ORA4MAS systems: and environmental resources. Related to organizations, some
(i) participating agents, assumed to join the organization in approaches propose agents able to automatically internalize or-
order to exploit its functions (i.e., adopting roles, committing ganizational specifications (i.e. MASQ, “normative objects”),
missions etc.), while (ii) organization aware agents, assumed to other approaches, as (ORA4MAS and SEI) assume agents’
manage the organization by making changes to its functional awareness to be encoded at a programming level.
and structural aspects (i.e., creating and updating functional
schemes or groups) or to make decisions about the deontic Finally, few approaches account technological integration,
events (i.e. norm violations). for instance with respect to varying agent architectures, proto-
cols and data types. Besides, the described proposals typically
B. Open Issues and Challenges focus on a restricted set of interaction styles (i.e. dialogical
Despite the richness of the models proposed for organiza- interactions supported by an institutional infrastructure in SEI,
tions of agents situated in computational environments, many environment mediated interactions in normative objects, an
aspects are still under discussion and have still to converge hybrid approach in ORA4MAS).
in a shared perspective between the different research lines. With the aim to respond the above mentioned challenges,
This variety of approaches have been dealt with separately the next sections describe an integrated approach aimed at
in current programming approaches, each forming a different devising a unified programming model seamlessly integrating
piece of a global view, with few consideration for how they agents, organizations and environments.
Visitor Staff
0..1
Escort Patient Doctor
0..1 1..1 1..1
Visit Staff
Group Group
0..NVMAX Surgery Room 1..1
Group
LEGEND
LINKS INTRA-GROUP EXTRA-GROUP
inheritance
min..max acquaintance
composition
communication
ROLE
authority
GROUP
ABS compatibility
ROLE
(a) Structural Specification (b) Deontic Specification
Fig. 1. Structural (a) and Normative (b) specifications for the hospital scenario, represented using the Moise graphical notation.
III. U NIFYING AGENTS , O RGANIZATIONS AND assumed to be played by a physician. It extends the properties
E NVIRONMENTS P ROGRAMMING of a more generic staff role, which is assigned in support and
This section figures out the main elements characterizing administration activities inside the group. Relationships can be
an Embodied Organization. It envisages an integrated MAS in specified between roles to define authorities, communication
terms of societies of agents, environmental and organizational channels and acquaintance links. Groups consist in a set of
entities. In doing this, we refer to the consistent body of roles and related properties and links. In the hospital scenario
work already addressed at specifying existing computational escorts and patients form visit groups, while staff and doctor
models, while only the aspects which are relevant for the from staff groups. The specification allows taxonomies of
purposes of this work will be detailed. In particular, we refer groups (i.e., escorts and patients forming visit group), and
to Jason [3] as agent development framework, CArtAgO [14] intra-group links, stating that an agent playing the source role
for environments and Moise [10] for organizations. is linked to all agents playing the target role. Notice that the
In order to ease the description, the approach will be cardinalities for roles inside a group are specified, indicating
sketched in the context of an hospital scenario. It summarizes the maximum amount of agents allowed to play that role.
the dynamics of an ambulatory room, and can be seen as an The constraints imposed by the SS allow to establish global
open system, where heterogenous agents can enter and leave in properties on groups, e.g. the well-formedness property means
order to fulfill their purposes. In particular, two types of agents to complain role cardinality, compatibility, and so on.
are modeled as organization participants. Staff agents (namely The Functional Specification (FS) gives a set of functional
physicians and medical nurses) are assumed to cooperate with schemes specifying how, according with the SS, various
each other in order to provide medical assistance to visitors. groups of agents are expected to achieve their global (orga-
Accordingly, visitor agents (namely patients and escorts) are nizational) goals. The related schemes can be seen as goal
assumed to interact themselves in order to book and exploit decomposition trees, where the root is a goal to be achieved
the medical examinations provided by the staff. by the overall group and the leafs are goals that can be
A. Organizations achieved by the single agents. A mission defines all the goals
an agent commits to when participating in the execution of
The first considered dimension concerns the organization.
a scheme and, accordingly, groups together coherent goals
We do adopt the Moise model, which allows to specify an
which are assigned to a role in a group. The FS for the
organization based on three different dimensions referred as (i)
hospital scenario (Fig. 2) presents three rehearsed schemes.
structural, (ii) functional, and (iii) normative1 . The Structural
The visitor scheme (visitorSch) describes the goal tree related
Specification (SS) provides the organizational structure in
to the visitor group. It specifies three missions, namely mVisit
terms of groups of agents, roles and functional relations
as the mission to which each agent joining the visit group has
between roles (links). A role defines the behavioral scope of
to commit, mPatient as the mission to be committed by the
agents actually playing it, thus providing a standardized pattern
patient who has to undergo the medical visit, and mPay as
of behavior for the autonomous part of the system. An inher-
the mission to be committed by at least one agent in the visit
itance relation can be specified, indicating roles that extend
group. Notice that the goals “do the visit” (which is related
and inherit properties from parent roles. As showed in Fig. 1
to the mission mPatient) and “pay visit” (which is related to
(left), visitor agents can adopt two roles, patient and escort,
the mission mPay) can be fulfilled in parallel. The monitorSch
both inheriting from a visitor abstract role. The doctor role is
describes the activities performed by a staff agent. These plans
1 We here provide a synthesis of the Moise approach showing the speci- are aimed at verifying if the activities performed by the visitors
fication of the hospital scenario. For a more detailed description, see [10]. follow an expected outcome, namely if the visitors fulfill the
visitorSch monitorSch docSch
visitor monitor Doctor
scheme scheme scheme
mVisit mVisit visit mVisit mStaff mDoc
enter book exit enforcement visit
observe
the room the visit patient
mPatient mPay
do the pay mRew mSan
visit visit send send
bill fee
use use
Desk BillingMachine focus use
bookVisit pay Desk, SurgeryTablet
use BillingMachine use use signDoc
joinWorkspace SurgeryTablet quitWorkspace Terminal Terminal
Hospital signPat Hospital sendBill sendFee
Hospital payments reservations
visits
ENVIRONMENT Workspace signDoc sendBill
MANAGEMENT pay bookVisit signPat sendFee
INFRASTRUCTURE
( EMI ) BillingMachine Desk SurgeryTablet Terminal
Fig. 2. (Above) Moise Functional Specification (FS) for the hospital scenario. Schemes are used to coordinate the behavior of autonomous agents. (Below)
FS is used to find a set of environmental artifacts, and to map their functionalities in the EMI.
payment committing the mPay mission (which includes the proach envisages organizational artifacts (OA) are those non-
“pay visit” goal). Finally, the docSch specifies the activities to autonomous computational entities adopted to reify organiza-
which a doctor has to commit, namely to perform the visit to tions at runtime, thereby implementing the institutional dimen-
every patient. Notice that each mission has a further property sion within the MAS. In particular, ORA4MAS adopts two
specifying the maximum amount of time than an agent has to types of artifacts, referred as scheme and group artifacts, which
commit to the mission (“time to fulfill”, or ttf value). The FS manage the organizational aspects as specified in Moise’s
also defines the expected cardinality for every mission in the functional, structural and normative dimensions. The result-
scheme, namely the number of agents inside the group who ing system has been referred as Organizational Management
may commit a given mission without violating the scheme Infrastructure (OMI), where the term infrastructure can be
constraints. understood from an agent perspective: it embeds those organi-
The Normative Specification (NS) relates roles (as they are zational functionalities exploitable by agents to participate the
specified in the SS) to missions (as they are specified in the FS) organizational activities and to access organization resources
by specifying a set of norms. Moise norms result in terms of possibly exploiting, creating and modifying OAs on the need.
permissions or obligations to commit to a mission. This allows Of course, in order to suitably exploit the OMI functionalities,
goals to be indirectly related to roles and groups, i.e. through agents need to be equipped with special capabilities and
the policies specified for mission commitment. Fig. 1 (right) knowledge about the organizational structures, that is what
shows the declarative specification of the norms regulating in Subsection II-B we refer as agent awareness.
the hospital scenario, and refers to the missions described in
Fig. 2. “Time to fulfill” (ttf ) values refer to the maximum B. Environments
amount of time the organization expects for the agent to fulfill As said in Subsection II-A, the ORA4MAS approach
a norm. For instance, norms n1 and n2 define an obligation does not support environments besides organizational func-
for agents playing either patient and escort roles to commit to tionalities. To this end, dually to the OMI, an Environment
the mVisit mission. A patient is further obliged to commit to Management Infrastructure (EMI) is introduced to embed the
mPatient mission (n3). The norm n10 is activated only when set of environmental entities aimed at supporting pragmatic
the norm n6 is not fulfilled: It specifies an obligation for a functionalities. While artifacts are adopted as basic building
doctor to commit the mStaff mission, if no other staff agent blocks to implement the EMI, environments also make use
is committing to it inside the group. Based on the constraints of workspaces (e.g., an Hospital workspace is assumed
specified within the SS and FS, the NS is assumed to include to contain the hospital infrastructures). Artifacts are adopted
an additional set of norms which are automatically generated in this case to provide a concrete (brute) dimension – at
in order to control role cardinality, goal compliance, deadline the environment level – to the global system. Workspace are
of commitments, etc. adopted in order to model a notion of locality in terms of an
The concrete computational entities based on the above application domain.
detailed specification have been developed based on an ex- As Fig. 2 shows, it is quite straightforward to find a basic
tended version of ORA4MAS [11]. This programming ap- set of Environment Artifacts (EA) building the EMI. Taking an
agent perspective, the developer here simply imagines which nature of artifact operations is assumed to cover the functional
kind of service may be required for the fulfillment of the dimension.
various missions/goals, thus mapping artifact functionalities
to the functional specification given by the Moise FS. IV. E MBODIED O RGANIZATIONS
Designing an EMI is thus not dissimilar to instrumenting a As far as the global system is conceived, EMI and OMI are
real workplace in the human case: (i) to model the hospital situated side by side inside the same work environment, but
room it will be used a specialized hospital workspace, (ii) to they are conceived as separated systems. They are assumed
automate bookings it will be provided a Desk artifact, (iii) to to face distinct application domains, the former being related
finalize visits it will be provided a (program running on an) to concrete environment functionalities and the latter dealing
Surgery Tablet artifact, (iv) to automate payments it will be specifically with organizational ones. The notion of Embodied
provided a Billing Machine artifact, and (v) to send fees and Organization provides a more strict integration: it further
bills it will be provided a Terminal artifact. identifies and implements additional mechanisms and con-
ceives a unified infrastructure enabling functional relationships
C. Agents between EMI and OMI. As some of the approaches discussed
in Section II, we theoretically found this relationship on
Besides the abstract indication of the different artifacts Searle’s notion of constitutive rules. Differently from other
exploitable at the environment level, the Fig. 2 also shows the approaches, we ground the notion of Embodied Organization
actions to be performed by agents for achieving their goals. on a concrete programming model, as the one who lead us to
Thanks to the CArtAgO integration technology, several agent the implementation of EMI and OMI. As explained below, Em-
platforms are actually enabled to play in environments: seam- bodied Organizations rely on a revised management of events
less interoperability is provided by implementing a basic set of in CArtAgO, and can be specified by special programming
actions, and related perception mechanisms, allowing agents constructs referred as Emb-Org-Rules.
to interact with artifacts and workspaces [14], [15]. Those
actions are directly mapped into artifact operations (functions), A. Events
or addressed to the workspace: in the case of the EMI, a Jason A crucial element characterizing Embodied Organizations
agent has to perform a joinWorkspace("Hospital") is given by the renewed workspace kernel based on events.
action to enter the room (which is related to the mVisit Events are records of significant changes in the application
mission); to book the visit (related to the mVisit mission) the domain, handled at a platform level inside CArtAgO. They are
action bookvisit()[artifact_name("Desk")] has referred to both state and processes to represent the transitions
to be performed on the desk artifact, and so on (see Fig. 2, of configurations inside workspaces. Each event is represented
below). by a type,value pair (hevt , evv i): Event type indicates the
The same semantic mapping agents’ actions into artifact type of the event (i.e., join_req indicating agents joining
operations is adopted to describe interactions between agents workspace, op_completed indicating the completion of an
and OMI: e.g., commitMission is an operation that can artifact operation, signal indicating events signalled within
be used by agents upon the scheme artifact to notify mission artifact operation execution, and so on); Event value gives
commitments; adoptRole (or leaveRole) can be used by additional information about the event (i.e., the source of
an agent upon the group artifact in order to adopt (leave) a the event, its informational content, and so on). Due to the
given role inside the group, etc. lack of space, the complete list of events, together with the
Fig. 3 (left) shows a global picture of the resulting system. description of the mechanism underlying event processing,
As showed, agents fulfill their goals and coordinate them- can not be described here. The interested reader can find the
selves by interacting with EMI artifacts, while staff agents, complete model, including the formal transition system, in
which we assume as special agents aware of organizational [13]. We here emphasize the relevance of events, which have
functionalities, can directly interact with the OMI. Both these the twofold role (i) to be perceived or triggered by agents (i.e.
dimensions are an integral part of the global infrastructure focusing/using artifacts) and (ii) to be collected and ranked
and, most important, can be dynamically exploited by agents within the workspace in order to trace the global dynamic of
to serve their purposes. From an agent perspective, the whole the system.
system can be understood as a set of facts and functions,
which are exploited, from time to time, to the organizational B. Embodied Organization Rules
and environmental dimensions. Through artifacts, the global While the former role played by events refers to the interac-
infrastructure provides observable states, namely information tion between agents and artifacts, the second role is exploited
readable by agents for improving their knowledge. Artifacts to identify, and possibly govern, intra-workspace dynamics.
also provide operations, namely process based functionalities, On such a basis, the notion of Embodied Organization refers
aimed at being exploited by agents for externalizing activities to the particular class of situated organization structured in
in terms of external actions. Thus, the epistemic nature of terms of artifact based infrastructures and governed by consti-
observable properties can be addressed to the informational tutive rules based on workspace events. Events are originated
dimension of the whole infrastructure, while the pragmatic within the infrastructure, being produced by environmental
Environment
Event
EMI Event
ENVIRONMENT
STAFF ARTIFACTS Ev Type
Ev Value Organization
STAFF Event
Terminal Triggers
GroupBoards SurgeryTablet
OMI Constitutive Rule 1..n Embodied
ORGANISATIONAL (Emb-Org-Rule) Organization
ARTIFACTS Produces
SchemeBoards
VISITOR
BillingMachine
Count-as Enact
Hospital Desk Rule Rule
VISITOR
Workspace
Agent
Platforms
Fig. 3. (Left) Global view of the system presents an open set of agents at work with infrastructures managing Environment and Organization. Functional
relationships between EMI and OMI are established by count-as and enact rules. (Right) Meta-model for Organizational Embodied Rules, used to implement
count-as and enact rules.
and organizational entities. Computing constitutive rules is visitor, which purpose is to book a medical visit and possibly
realized by Emb-Org-Rule, which consist of a programmable achieve it. Thus, an event join req, hvaid , ti, dispatched once
constructs “gluing” together organizational and environmental an agent vaid tries to enter the workspace, from the point
dimensions. An abstract model of this process is shown by of view of the organization “count-as” creating a new position
the dotted arrows between EMI and OMI in Fig. 3 (right). related to the visit group. Making the event join req to “count
Structures defining Emb-Org-Rule refer to count-as and enact as” vaid adopting the role visitor, is specified by the first
relations. rule in TABLE I (left): it states that since an event signalling
Count-as rules state which are the consequences, at the that an agent Ag is joining the workspace, an Emb-Org-
organizational level, for an event generated inside the overall Rule must be applied to the system. The body of the rule
infrastructure. They indicate how, since the actions performed specifies that two new instances of organizational artifacts
by the agents, the system automatically detects relevant events, related to the visit group will be created using the make
thus transforming them to the application of a set of operators operator. In this case the new artifacts will be identified
aimed at changing the configuration of the Embodied Orga- by visitorGroupBoard and visitorSchBoard. The
nization. In so doing, either relevant events occurring inside following operator constitutes the new role inside the group:
the EMI (possibly triggered by agents actions), either events apply acts on the visitorGroupBoard artifact just created
occurring in the context of the organization itself (OMI) can by automatically making the agent Ag to adopt the role patient.
be vehicled to the institutional dimension: these events can be Finally, once the adopt role operator succeeds, the last operator
further translated in the opportune institutional changes inside includes the agent Ag in the workspace.
the OMI, that is assumed to update accordingly. In the above described scenario, the effect of the
Enact rules state, for each institutional event, which is the application of the rule provides an institutional out-
control feedback at the environmental level. Hence, enact come to the joinWorkspace actions. Besides joining the
rules express how the organizational entities automatically workspace, a sequence of operators is applied establish-
control the environmental ones. The use of enact rules allows ing what this event means in organizational terms. When
to exploit organizational events (i.e. role adoption, mission the effects of the role-adoption are committed, as previ-
commitment) in order to elicit changes in the environment. ously described, a new event is generated by the group
board: hop completed, h"visitorGroupBoard", vaid ,
V. P ROGRAMMING E MBODIED O RGANIZATIONS adoptRole, patient ii. For the organization, such an
Embodied Organizations enable a unified perspective on event may “count-as” committing to mission mP at on the
agents, organizations and environments by conceiving an in- visitorSchBoard. This relation is specified by the second
teraction space based on a twofold infrastructure governed by rule in TABLE I, where a commitMission is applied to
events and constitutive rules (Emb-Org-Rules). In this section the visitorSchBoard for the mission mPat. Similarly,
examples of programming such rules are discussed. an event hws leaved, hvaid , tii, signalling that the visitor
Programming Count-as Rules According to the Moise FS agent has left the workspace, from an organizational per-
previously defined, the organization expects that an agent vaid spective “count-as” leaving the role patient. This relation
joining the hospital workspace is assumed to play the role is specified by the first rule in TABLE I (right), where
+join_req(Ag)
+ws_leaved(Ag)
-> make("visitorGroupBoard",
-> apply("visitorGroupBoard",
"OMI.GroupBoard",
leaveRole(Ag, "patient")).
["moise/hospital.xml","visitGroup"]);
make("visitorSchBoard",
+op_completed("BillingMachine",
"OMI.SchemeBoard",
Ag, pay)
["moise/hospital.xml","visitorSch"]);
-> apply("visitorSchBoard",
apply("visitorGroupBoard",
setGoalAchieved(Ag, pay_visit)).
adoptRole(Ag, "patient"));
include(Ag).
+op_completed("Terminal",
+op_completed("visitorGroupBoard", _,
Ag, sendFee)
adoptRole(Ag, "patient"))
-> apply("monitorSchBoard",
-> apply("visitorSchBoard",
setGoalAchieved(Ag, send_fee)).
commitMission(Ag, "mPat")).
TABLE I
E XAMPLE OF E MB -O RG -RULE ( COUNT- AS ) IN THE HOSPITAL SCENARIO .
+signal("monitorSchBoard",
goal_non_compliance,
+signal("visitorGroupBoard",
obligation(Ag,
role_cardinality, visitor)
ngoa(monitorSch,mRew,send_bill),
-> disable("Desk", bookVisit).
achieved(monitorSch,send_bill,Ag), TTF)
-> exclude(Ag).
TABLE II
E XAMPLE OF E MB -O RG -RULE ( ENACT ) IN THE HOSPITAL SCENARIO .
a leaveRole is applied to the visitorGroupBoard scheme board assigned to the monitorSch to signal the event
for the role patient. At the same time, an event like hsignal, hmonitorSchBoard, goal_non_compliance,
hop completed, hBillingMachine, vaid , pay, tii signals obligation(Ag,ngoa(monitorSch,mRew,send_bill),
that a visitor agent has successfully finalized the pay operation achieved(monitorSch,send_bill,Ag),TTF), tii.
upon the billing machine. Such an event “count-as” having This event is generated thanks to a special norm (called
achieved the goal pay visit on the visitorSchBoard (sec- goal_non_compliance) which is automatically generated
ond rule in TABLE I, right). Finally, an event hop completed, since the Moise specification and stored inside the OMI.
hTerminal, said , sendFee , tii, signalling that a staff agent Due to the enact rule specified in TABLE II (right), this
has successfully used the terminal to send the fee to a given causes the exclusion for the Ag agent from the hospital
patient, “count-as” having achieved the goal send fee (third workspace.
rule in TABLE I, right).
Programming Enact Rules Enact effects are defined to indi- VI. C ONCLUSION AND P ERSPECTIVES
cate how, from the events occurring at the institutional level, The notion of Embodied Organization has been introduced
some control feedback can be applied to the environmental as a unified programming model for a seamless integration of
infrastructure. As far as the execution of the operations is environmental and organizational dimensions of MAS.
conceived in CArtAgO, the OMI automatically dispatches In Embodied Organizations, either environmental and or-
events signalling ongoing violations. Violations are thus or- ganizational entities are implemented in concrete infrastruc-
ganizational events which may suddenly elicit the application tures instrumenting workspaces, decentralized in specialized
of some enact rule used to regiment the environment. artifacts which serve informational and operational functions.
In TABLE II, a regimentation is installed by the The approach establishes a coherent semantic for agent - in-
organization thanks to the enact rule stating that an event frastructure interactions, Embodied Organizations define func-
hsignal, hvisitorGroupBoard, role_cardinality, tional relationships between the heterogenous entities at the
∅, tii signalled by the visitorGroupBoard indicates basis of organizations and environments. These are placed in
the violation for the norm role_ cardinality. The terms of programmable constructs (Emb-Org-Rules), gov-
related enact rule is given in TABLE II (left), where the erned by workspace events and inspired by Searle’s notion
reaction to this event is specified in order to disable the of constitutive rules. Implementing organizations in concrete
book operation on the desk artifact, for all the agents environments allows to deal explicitly with goals and norms,
inside the workspace. The absence of any parameter which fulfillment can be structurally monitored and promoted
related to agent identifier in the disable("Desk", at the organizational level through the use of artifacts. Em-
bookVisit) operator makes the disabling to affect bodied Organizations are aimed to fit the work of agents and
the overall set of agents inside the workspace. Similarly, accordingly to allow them to externalize pragmatic and organi-
violating the obligation imposed to the staff agent to zational activities. The use of Emb-Org-Rule automates and
fulfill sanctioning and rewarding missions elicits the promotes specific organizational patterns, to which agents may
effortlessly participate simply by exploiting environmental [14] Alessandro Ricci, Michele Piunti, and Mirko Viroli. Environment
resources. Artifacts can be used in goal oriented activities, and, programming in multi-agent systems: An artifact-based perspective.
Autonomous Agents and Multi-Agent Systems, 2010. Springer, ISSN
most important, without the need to be aware of organizational 1387-2532 (Print) 1573-7454 (Online).
notions like roles, norms, etc. Technological interoperability is [15] Alessandro Ricci, Andrea Santi, and Michele Piunti. Action and
ensured at a system level, by providing mechanisms for agent- Perception in Multi-Agent Programming Languages: From Exogenous
to Endogenous Environments. In Proceedings Programming Multiagent
artifact interactions which are based on a coherent semantic Systems (PROMAS-10), 2010.
defined in CArtAgO. Besides, several interaction styles can be [16] John R. Searle. Speech Acts, chapter What is a Speech Act? Cambridge
established at an application level, being agents mediated by University Press, 1964.
[17] John R. Searle. The Construction of Social Reality. Free Press, 1997.
infrastructures which can be modified, replaced and created [18] M. Sierhuis. Modeling and Simulating Work Practice; Brahms: A
on the need. multiagent modeling and simulation language for work system analysis
Future work will be addressed at covering missing aspects, and design. PhD thesis, University of Amsterdam, SIKS Dissertation
Series, 2001.
such as the dialogical dimension of interactions, and the [19] Tiberiu Stratulat, Jacques Ferber, and John Tranier. MASQ: Towards an
inclusion of real embodied entities in the system (i.e., humans, Integral Approach of Agent-Based Interaction. In Proc. of 8th Conf. on
robots, etc.). An important objective is the definition of a Agents and Multi Agent Systems (AAMAS-09), 2009.
[20] Nick Tinnemeier, Mehdi Dastani, J.-J.Ch. Meyer, and L. van der
general purpose approach, towards the full adoption of the Torre. Programming normative artifacts with declarative obligations and
proposed model in the context of concrete application domains prohibitions. In IEEE/WIC/ACM International Joint Conference on Web
and mainstream agent oriented programming. Intelligence and Intelligent Agent Technology (WI-IAT 2009), 2009.
[21] M. Birna van Riemsdijk, Koen Hindriks, and Catholijin Jonker. Pro-
gramming organisation-aware agents: a research agenda. In In 10th
R EFERENCES Engineering Societies for Agents Worlds (ESAW 09), 2009.
[1] José-Antonio Báez-Barranco, Tiberiu Stratulat, and Jacques Ferber. A [22] Danny Weyns, Andrea Omicini, and James J. Odell. Environment as
unified model for physical and social environments. In Environments for a first-class abstraction in multi-agent systems. Autonomous Agents
Multi-Agent Systems III, Third International Workshop (E4MAS 2006), and Multi-Agent Systems, 14(1):5–30, February 2007. Special Issue on
volume 4389 of Lecture Notes in Computer Science, pages 41–50. Environments for Multi-agent Systems.
Springer, 2006.
[2] Olivier Boissier, Jomi Fred Hübner, and Jaime Simão Sichman. Orga-
nization Oriented Programming: From Closed to Open Organizations.
In Engineering Societies for Agent Worlds (ESAW-2006). Extended and
Revised version in Lecture Notes in Computer Science LNCS series,
Springer, pages 86–105, 2006.
[3] Rafael H. Bordini, Jomi Fred Hübner, and Michael Wooldrige. Program-
ming Multi-Agent Systems in AgentSpeak using Jason. Wiley Series in
Agent Technology. John Wiley & Sons, 2007.
[4] Jordi Campos, Maite Lòopez-Sànchez, Juan A. Rodrı̀guez-Aguilar, and
Marc Esteva. Formalising Situatedness and Adaptation in Electronic
Institutions. In COIN-08, Proc., 2008.
[5] Mehdi Dastani, Nick Tinnemeier, and John-Jules CH. Meyer. A
programming language for normative multi-agent systems. In Multi-
Agent Systems: Semantics and Dynamics of Organizational Models. IGI-
Global, 2009.
[6] Virginia Dignum, editor. Handbook of Research on Multi-Agent Systems:
Semantics and Dynamics of Organizational Models. IGI-Global, 2009.
[7] Marc Esteva, Juan A. Rodrı́guez-Aguilar, Bruno Rosell, and Josep L.
AMELI: An agent-based middleware for electronic institutions. In
Proceedings of International conference on Autonomous Agents and
Multi Agent Systems (AAMAS’04), pages 236–243, New York, 2004.
ACM.
[8] Jacques Ferber, Olivier Gutknecht, and Fabien Michel. From Agents
to Organizations: An Organizational View of Multi-agent Systems. In
Proceedings of (AOSE-03), volume 2935 of Lecture Notes Computer
Science (LNCS). Springer, 2003.
[9] Jacques Ferber and Jean-Pierre Müller. Influences and Reaction: a Model
of Situated Multi-Agent Systems. In Proc. of the 2nd Int. Conf. on
Multi-Agent Systems (ICMAS’96). AAAI, 1996.
[10] Jomi F. Hübner, , Jaime S. Sichman, and Olivier Boissier. Developing
organised multi-agent systems using the MOISE+ model: Programming
issues at the system and agent levels. International Journal of Agent-
Oriented Software Engineering, 1(3/4):370–395, 2007.
[11] Jomi F. Hübner, Olivier Boissier, Rosine Kitio, and Alessandro Ricci.
Instrumenting Multi-Agent Organisations with Organisational Artifacts
and Agents. Journal of Autonomous Agents and Multi-Agent Systems,
April 2009.
[12] Fabio Y. Okuyama, Rafael H. Bordini, and Antônio Carlos
da Rocha Costa. A Distributed Normative Infrastructure for Situated
Multi-Agent Organisations. In Decl. Agent Lang. & Techn. (DALT-VI),
volume 5397 of LNCS. Springer, 2009.
[13] Michele Piunti. Designing and Programming Organizational Infrastruc-
tures for Agents situated in Artifact-based Environments. PhD thesis,
ALMA MATER STUDIORUM Universitá di Bologna, April 2010.