We present the Tokyo Virtual Living Lab, an experimental space for conducting controlled driving and travel studies. The virtual lab is based on 3D Internet technology and supports studies involving multiple users in the same shared space. This feature is crucial for analyzing interactive driving behaviors in future smart cities. The novelty of this work is two-fold. First, we developed a pipeline that takes free map data as input and outputs a semantically enriched graphical structure, called Navigation Network. Second, we built a Navigation Segment Agent that coordinates a Multiagent traffic simulator based on the Navigation Network and supports the integration of user controlled vehicles. This approach can significantly reduce the effort of preparing realistic driving behavior studies. To demonstrate the usefulness of our approach, we created a 3D model of a part of Tokyo and performed an experiment with human drivers in two conditions: ubiquitous eco-traffic and "normal" traffic