Research Article

Agent-Oriented Software Engineering Methodologies: Analysis and Future Directions

Table 3

Comparative analysis of ADELFE, ASEME, and Prometheus methodologies.

MethodologyStrengthsLimitationsApplication domain

ADELFE(i) Can be used by a nonspecialist in agent systems(i) It is specialized; thus, it cannot be used to design all existing applications or model all kinds of agents(i) Tools for self-organizing applications engineering [49]
(ii) Will also be easier to combine the parts from other approaches into ADELFE(ii) Some definitions of work still do not exist(ii) A sample application of ADELFE focusing on analysis and design of the mechanical synthesis problem [43]
(iii) Use the cooperation rules [49](iii) Sometimes the developer may find the graphical modeling tool difficult to use [49](iii) A tool for generating model transformations by example in MASs [62]
(iv) Allows different UML/AUML realizations(iv) ADELFE 3.0 design, building adaptive multiagent systems based on simulation a case study [1]

ASEME(i) Supports documentation of nonfunctional requirements(i) The guidelines are missing because the authors rely only on the paradigm shifts(i) Automated product pricing using argumentation [54]
(ii) Provides the model transformations among the various development stages(ii) The same thing happens with the tasks to be conducted in each stage of development(ii) Using ASEME methodology for model-driven agent systems development [35]
(iii) Some processes need to be improved(iii) Engineering an agent-based system for product pricing and automation [35]
(iv) In requirements, implementation, and design, there are no guidelines for the production of models(iv) Using agent-based methodologies in healthcare information systems [47]

Prometheus(i) Offers a complete lifecycle methodology from requirements specification to detailed design(i) There is less focus on initial requirements(i) An open meteorological alerting system: issues and solutions [63]
(ii) Supports both dynamic and static models for individual agents(ii) It does not deal at all with mobile agents(ii) Tool support for agent development using the Prometheus methodology in quality software [64]
(iii) Provides elaborated guidance on how to carry out the different stages(iii) It cannot be established on UML(iii) AO modeling and development of a person-following mobile robot [65]
(iv) Gives clear support to the environment concept(iv) Support for the socialite side of the agent focuses on the lowest common divisors: messages and protocols