Researchs in the team are definitely in a multidisciplinary approach and seek to develop “virtual test tubes” promoting the study of complex systems through observation of autonomous entities using artificial intelligence behaviours. This approach is called “individual-based” because its goal is to obtain the expected group phenomena by aggregation of individual behaviors, in other words to allow the passage of the microscopic level observation (individuals) to the macroscopic levels (groups). This approach is particularly suitable for the realization of simulators in such diverse fields as computational finance, crowd simulation, traffic or cell biology. In these areas, tools and platforms created by the team are used as well as for simulation than for Serious Game or problem solving. The team works in particular on the development and coding of intelligent behaviors for artificial entities, development of software platforms to exploit them, and development of methods to go from a specific problem analysis to its concrete realization. The team is very sensitive to practical and concrete aspects of this approach. It offers notably several platforms for studiying these problems with the highest level of effectiveness and with a constant concern to avoid bias in the simulations.