Contemporary cities are, in essence, Complex Systems (CS) extremely interesting to investigate. A key feature is that Urban Environments provide big databases which are useful to analyze and to understand partial views of their own structure and dynamics. This represents an emergent research field where the generation and use of physical and digital information from cities is applied to design new applications, services and urban plans, as well as, to analysis social behaviors in cities. This field is multidisciplinary, in nature, where Computer Science, Urban Studies, e-Government, media studies, Open Data strategies and Web Services play important roles. Digital information about cities have different reliability. Current initiatives based on OpenData (e.g. ParisData, http://opendata.paris.fr/) aim to process this information, hoping to build useful mashups for cities.
The aim of this project is to describe our experience analyzing WWW Data and Web Services, which are needed to simulate urban phenomenas, specifically urban catastrophes, and it is based on every modern city has a digital mirror where researches can access to information (and knowledge) about the city. The analysis have been developed into the context of «eCompleXcity project». One goal of this project is to study Complex Systems associated to Urban Informatics and Computing.
We are interested in simulating urban processes by means of Multiagent Systems (MAS). MAS point of view of CS is quite natural because it considers that each module/node of system is an agent. From the Agent point of view, we can formally define its own behavior, as well as, its interaction with other agents. CS can evolve along time and new properties and behaviors emerge. Powerful tools to design CS in a high abstraction level are available. These tools also let us to create big scalable simulations(experiments) and this means a key point for MAS.
Our framework has been applied to a case study: simulation of flooding caused by Katrina’s hurricane in New Orleans.