Simulation in EA

General Objective:

To develop a platform that allows to model enterprise architectures in order to analyze its dynamic aspects and supports evaluation of alternative designs to help the decision making process in EA projects.

 Specific Objectives:

  • To develop a model-driven platform to simulate enterprise architectures based in a discrete events simulation approximation.
  • To define languages to express internal and external factors that affect the organization's reactions to different events in time, in order to enrich the EA's simulation.
  • To define a language to express business indicators in order to allow collecting data about the simulation to evaluate non-functional aspects of the EA model.
  • To define a language to express the configuration of the model to test different possibilities of the enterprise architecture.
  • To develop tools with the purpose of allowing the definition of the necessary simulation artifacts.
  • To develop a tool that allows presenting the simulation results as a dashboard, according to the representation specified in the business indicators.
  • To develop composition mechanisms for simulation domain metamodels.
  • To develop composition mechanisms for domain indicators.

To develop a repository to store domain metamodels and indicators, and composition definitions.

Detailed information:

  • The platform runs on Cumbia. Metamodels and models are defined using Cumbia’s XML definition.
  • EA concepts are materialized using OpenObjects. The behavior of these concepts is defined using state machines.
  • The specification of indicators and stimuli is done using two DSLs designed.
  • The specification of configuration is done using a properties file.
  • Metamodels, models and indicators are stored in repositories.
  • The simulation is managed by a controller, who uses a timeline to coordinate chronology between events.
  • The results of the simulation are gathered by the observation model, which is composed by a series of sensors placed in elements’ state machines and traces that store traces.

Results of the simulation are presented according to the domain indicators specified in a dashboard.

Simulation in EA


 Laura Manzur




 In progress

  • Design of adapters to consume the engine from other modeling platforms such as ArchiMate
  • Semi-automatic import of files from other languages for modeling Enterprise Architecture to the engine.
  • Development of graphical editors for the used DSLs.
  • Import of metamodels, models and indicators from repository.
  • Define ideal domain metamodels and indicators for different business verticals.

  • Laura Manzur (Active)
  • John Santa (Inactive)
  • Jorge Mario Ulloa (Active)