Deadline: 
13 Febrero 2007
Institution: 
University of Klagenfurt (Austria) / IIIA - CSIC
Speaker: 
Ernst Gebetsroither

During the past 10 years, two approaches have been often used to simulate processes within complex socio-economic and physical systems: The System Dynamics (SD) approach developed in the 1960?s by J.W. Forrester and the Multi-Agent-Based Modelling (MAS) approach developed during the 1990?s. (by e.g. the Santa Fe Institute). Both enable to simulate non-linear, feedback driven complex dynamic systems. ??????? The SD approach looks at the system from a macro scale -- a top down point of view -- and enables to simulate system behaviour as a result of an analysis of the general system structure and the integrated response as causal loop depicting the total stocks and flows between. (Forrester 19611, Meadows, et al.19922). This method is useful to simulate the development of the entire system considering certain side conditions change over time. However, in many cases the SD approach is insufficient, because it does not consider inherent self-organization processes (and interventions) at a micro scale. ??????? The MAS approach looks at a system from a micro scale. It defines a number of system elements (?agents?) that decide and behave individually aiming in certain spatial patterns relating from certain behaviour rules which may change steadily due to changes within the system. A MAS is a collection of autonomous agents acting independently using mainly local information and with, possibly, the ability to communicate with each other. Thus ABM can be used to explain emerging patterns of system behaviour. This structure enhances significantly the simulation of the system?s response on certain stages of transition and intervention on a very local scale. Within the last few years several research teams have been working on combining both methods in order to simulate the development of complex systems. The joining of SD and MAS to establish an integrated model is a promising approach for the development of decision support systems (DSS) for policy making on different levels and spatial scales. Such DSS?s will support the examination of decisions and response of different key players (stakeholders & decision makers) and their affect and feedback from actors at a micro level. Top down decisions in the past often failed, as they where neglecting self-organizing processes within the system. The understanding of feedback loops in complex socio-economic systems will not only enable to find better solutions for complex problems it will further help to achieve a higher acceptance for the solutions, during participatory processes.