Multiagent systems have become popular over the last few years for building complex, adaptive systems in a distributed, heterogeneous setting. Multiagent systems tend to be more robust and, in many cases, more efficient than single monolithic applications. However, unpredictable application environments make multiagent systems susceptible to individual failures that can significantly reduce its ability to accomplish its overall goal. The problem is that multiagent systems are typically designed to work within a limited set of configurations. Even when the system possesses the resources and computational power to accomplish its goal, it may be constrained by its own structure and knowledge of its member’s capabilities. To overcome these problems, we  are developing a framework that allows  the system to design its own organization at runtime.  This paper presents a key component of that framework, a metamodel for multiagent organizations named the Organization Model for Adaptive Computational Systems. This model defines the requisite knowledge of a system’s organizational structure and capabilities that will allow it to reorganize at runtime and enable it to achieve its goals effectively in the face of a changing environment and its agent’s capabilities.

While most people have an intuitive idea of what an organization is, when asked to define it explicitly, there are large numbers of “correct” answers. From early organizational research we learn that organizations have typically been defined as including the concepts of set of agents who play roles within a structure that defines the  relationships between the various roles [2]. Thus, we lay the foundation for our model by defining what is meant by goals (G), roles (R), and agents (A). We also add four additional entities: capabilities (C), assignments (Φ), policies (P), and a domain model (Σ). Capabilities are central to the process of determining which agents can play which roles and how well they can play them, while policies constrain the assignment of agents to roles thus controlling the allowable states of the organization. The domain model is a critical component that defines the ontology used to define behavioral policies and to allow agents to communicate effectively. 

OMACS defines an organization as a tuple O=<G, R, A, C, Φ, P, Σ,  oaf,  achieves,  capable,  require, possesses, potential> where 

• G  goals of the organization
• R   set of roles
• A   set of agents 
• C   set of capabilities
• Φ   relation over G  × R  × A that defines the current set of agent/role/goal assignments
• P   set of constraints on Φ 
• Σ   domain model used to specify objects in the environment, their inter-relationships, and the operations that can be performed upon them
• oaf   function Ρ(G × R × A) → [0.. ∞] that defines the quality of a proposed set of assignments
• achieves function G × R → [0..1] that defines how well a role achieves a goal
• capable function A × R → [0..1] that defines how well an agent can play a role
• requires   function R → Ρ(C) that defines the set of capabilities required to play a role
• possesses   function A × C → [0..1] that defines the quality of an agent’s capability 
• potential   function A × R × G → [0..1] that defines how well an agent can play a role to achieve a goal

Related Papers

• Scott A. DeLoach, Walamitien Oyenan & Eric T. Matson. A Capabilities Based Model for Artificial Organizations. Journal of Autonomous Agents and Multiagent Systems. Volume 16, no. 1, February 2008, pp. 13-56. DOI: 10.1007/s10458-007-9019-4.