Organizational adaptation to

CoRR 01/2011; abs/1110.4296. DOI: 10.1023/A:1018963630536
Source: DBLP

ABSTRACT A computational model of organizational adaptation in which change occurs at both the strategic and the operational level is presented. In this model, simulated annealing is used to alter the organization's structure even as the agents within the organization learn. Using this model a virtual experiment is run to generate hypotheses which can be tested in multiple venues. The results suggest that, although it may not be possible for organizations of complex adaptive agents to locate the optimal form, they can improve their performance by altering their structure. Moreover, organizations that most successfully adapt over time come to be larger, less dense, with fewer isolated agents, and fewer overlooked decision factors. These results have implications for organizations of both humans and non-humans. For example, they suggest that organizational learning resides not just in the minds of the personnel within the organization, but in the connections among personnel, and among personnel and tasks. These results suggest that collections of non-humans may come to seem more intelligent (i.e., show improved performance) even if the agents remain unchanged if the system simply develops duplicate copies of some of the artificial agents and if the connections among agents are dynamically altered.

  • [Show abstract] [Hide abstract]
    ABSTRACT: In this chapter, we discuss the design of adaptive hierarchical organizations for multi-agent systems (MAS). Hierarchical organizations have a number of advantages such as their ability to handle complex problems and their scalability to large organizations. By introducing adaptivity in the structure of hierarchical MAS organizations, we enable agents to balance resources in their organization. We will first provide a number of generic principles for the design of hierarchical MAS organizations. We show how these principles are used to design three different hierarchical organizations for a search and rescue task in the RoboCupRescue simulation environment. The first two of these organizations are static, and the third is able to adapt its structure. An empirical study on the performance of these three organizations shows that the dynamic organization performs better than the two static organizations.
    03/2010: pages 375-400;
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we present requirements for a successful learning approach for large scale multi-agent swarms: individual performance, cooperation and robustness. We then present a developmental, evolutionary approach for evolving hierarchical control structures for large (100-1000 agent), multi-agent swarms that addresses these requirements. Although hierarchical, the control structure does not suffer from single point of failures as do many hierarchical structures. The approach is tested on a novel problem for which a fully distributed swarm performs poorly. The results show that for some problems using an evolved control hierarchy to guide the agents leads to significantly better performance and scaling properties than fully distributed swarms using standard, simple behavioral rules. This research suggests that hierarchies are an important organizing feature to be considered for large multiagent tasks and that the efficient automated discovery of a hierarchy is an important research objective.
    Genetic and Evolutionary Computation Conference, GECCO 2010, Proceedings, Portland, Oregon, USA, July 7-11, 2010, Companion Material; 01/2010
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Three different strategies are identified for organizational adaptation, including dynamic process selection. An executable organizational model composed of individual models of a five stage interacting decision maker is used to evaluate the effectiveness of the different adaptation strategies on organizational performance. The concept of entropy is used to calculate the total activity value, a surrogate for decision maker workload, based on the functional partition and the adaptation strategy being implemented. The individual decision makers total activity is monitored, as overloaded decision makers constrain organizational performance. A virtual experiment was conducted; organizations implementing local and global adaptation strategies were compared to a control organization with no adaptation. The level of tolerance of the organization, the workload limit based on the concept of the bounded rationality constraint, was used to determined when a decision maker was overloaded: the limiting effect of the workload on performance. The timeliness of the organizations response was used in order to evaluate organizational output as a function of adaptation strategy.