Agent Oriented Software Engineering The MultiAgent approach offers a methodological framework well adapted to the analysis and modeling of complex systems. This approach considers systems as societies composed of autonomous and independent entities, called agents, that interact in order to solve problems or to achieve a common task. MultiAgent Systems (MAS) have been successfully applied to a great number of domains including robotics, distributed problem solving, modeling and simulation of complex systems, etc. In despite of the numerous applications, MAS exhibit a certain delay on formalization and modeling methodologies. Indeed, the design of this type of systems is often accomplished following an empiric procedure or on an ad hoc manner. The development of models and methodologies for this paradigm are of vital interest to the proper adoption by the scientists and industry. ASPECS - Agent-oriented Software Process for Engineering Complex Systems Our team colaborates in the definition of the ASPECS methodology, covering the analysis, design and implementation of (holonic) multi-agent systems (H/MAS). This methodology is based upon an holonic organisational metamodel and provides a step-by-step guide from requirements to code allowing the modelling of a system at different levels of details using a set of refinement methods. It integrates design models and philosophies from both object- and agent-oriented software engineering (OOSE and AOSE). It has been built by adopting the Model Driven Architecture (MDA) and thus we defined three levels of models. The main concepts of the organizational metamodels are roles, interactions and organizations. These concepts allow for abstraction and decomposition of MAS and HMAS. ASPECS uses UML as modelling language, the UML semantics and notation are used as reference points, but they have been extended by introducing new specific profiles to fulfil the specific needs of agents and holonic organisational design. The Janus_Platform provides an implementation and execution environment for models designed with ASPECS. Multi-Agent Systems formal specification and verification Formal methods provide several benefits to Software Engineering. As systems become more complex and safety becomes a priority in system design, formal methods offer mechanisms and tools for system verification. Providing formal specifications and methods for MAS meta-models and methodologies is imperious for the adoption of this technology, especially when MAS control critical environments such as nuclear plants, aerospace systems, etc. Optimization Genetic Algorithms Genetic Algorithms (GAs) are a family of computational models that simulate Darwin’s principle of natural selection in order to solve complex optimization problems (NP-Hard). They operate in parallel on a population of solution candidates, using the principle of survival of the fittest to successively produce better solutions. Genetic Algorithms are well suited to solve complex optimization problems that are compicated for traditional optimization approaches. They have been succesfully applied to automated design, robotics, vehicle routing, medicine, biology, aviation, telecomunications, etc. Despite the fact that they have been in use for a long time, the efficient design of real-life application is still a challenging task that requires knowledge and experience, both in the Genetic Algorithm and on the problem. The research and development around the methodologies required to design real life applications of Genetic Algorithms is very important at the time of solving real life problems coming from different areas of Science and Technology. Neural Networks Artificial Neural Networks is a family of computational models inspired by the human nervous system. They use many simple processing units (neurons), that work in parallel to achieve complex behaviors. Their mainThey have been succesfully applied to a number of fields. They are best suited for problems that are difficult for traditional algorithmic approaches, where a mathematical model is not available (model free estimators). This allows its use in Pattern Recognition problems, prediction, classification, optimization and control, without the need of a mathematical model. Neural Networks have successfully been applied in proces control, weather forecast, chemistry, information analysis, medicine, biology, concrete and soil technology and materilals science and technology, etc. Application of Neural Networks to new areas or problems is a challenging task, which is the subject of active research around the world.