Printer friendly version
Intelligent systems seek to simulate the migration processes
09 September 2011
Group Information Technology and Artificial Intelligence (GTI-IA) of the Polytechnic University of Valencia and the Boeing Research and Technology Europe (BR & TE), are carrying out a project for use in the management and analysis of phenomena present in migration.
The objective of this project is the simulation of social behavior of people within a specific scenario. Because the migration process involves a complex system of institutions that are influenced by many variables, this scenario is ideal to validate the project's objectives, although it could be applied in other areas such as mass evacuation, large design spaces as airports and shopping malls, etc.. Thanks to the simulation of behavior can be anticipated disaster solutions, economic crises, political or health.
Currently research is in the process of technology validation, have been implemented not only the decision-making models of the agents, but have also been investigated and incorporated technologies to evaluate the performance synthetic labor and financial markets as a key in the phenomenon.
Thus, researchers at the UPV have implemented the concepts of artificial economy in order to represent the influence of labor and financial markets over the flow of people on certain migratory routes. "We have also simulated the addition of security systems and port control stations through the use of radar, and we have analyzed and implemented a social network model that includes the type of participants, the level of relationship / influence between them and the type information is perceived, "said Vicente Botti, a researcher at Group Information Technology and Artificial Intelligence at the UPV.
This year the collaboration between BR & TE and GTI-IA has focused on the research of technologies for simulation environments using specific scenarios. At first he worked in immigration from West Africa to the Canary Islands and Cadiz, and now the scenario has been extended to include immigration from Nigeria and Senegal to Ceuta and Melilla.
This work is part of the project Research in Technologies for the Management of Immigration (Integra), funded by the Center for Industrial Technological Development, under the Cenit program of the Ministry of Science and Innovation. This project, carried out by a consortium of 15 Spanish companies and 25 research organizations, began in 2008 and is scheduled to end in 2012.
What are agents and multiagent systems
A complex system can be defined as any system that incorporates a large number of interconnected components, whose links contain additional information and hidden from the observer. As a result of interactions between components, new properties emerge that can not be explained by the properties of individual components. These properties are called emergent properties. Are examples of complex systems, ant colonies, economies and human social structures, climate, nervous systems, cells and living organisms, including humans. An intelligent agent is a computational process can perform tasks autonomously and communicates with other agents to solve problems through cooperation, coordination and negotiation. And a multi-agent system is a system composed of multiple interacting intelligent agents. These agents can cooperate or compete (or a combination of cooperation and competition) to achieve individual or collective tasks.
Multi-agent systems have proven to be a contemporary approach that can be used to study (using modeling techniques) complex systems, mainly because they incorporate the interactions between agents and the properties that emerge at higher levels, thus helping to understand the behavior of systems.
Various studies have been developed to model patterns of behavior within the field of multi-agent systems.
"Among the most important for our study are the study and analysis of patterns of human behavior and different animal species. In the area of animal behavior there are several works focused on the study of patterns of behavior, such as ant colonies, bees, shoals of fish, or migratory behavior of birds, among others. For the study of migration of animals, you can see patterns where their movements are predictable characteristics, seasons and years, involving the displacement of large numbers of individuals from one place to another. However, in the case of migration of people, this is influenced by political, economic, social and demographic changes at the level of complexity increases considerably. Therefore the predictive techniques used are also more complex, "says Vicente Botti.
Thus, according to the researchers, the agent technology to support decision making during the migration process is presented as a viable technology as it contributes to: define complex behaviors of each of the entities involved in each scenarios, model the processes and activities at individual and organizational units in each stage or simulate the behavior of the models, among other advantages.
The UPV researchers