Product engineering activities such as design, analysis, manufacturing planning, and control of equipment rely upon computer models. Descriptions of engineering objects such as elements and structures of products, results of tests and analyses, and engineering processes for design, analysis, and manufacturing of products have developed into very complex integrated data structures. Application of these complex product models has been extended to support lifecycle management of product data. Engineering utilizes advanced characteristics of modeling software such as analysis of behaviors of modeled objects, adaptive actions for controlled modifications of engineering objects, selection of optimal sets of parameters, etc. The next step of integration is awaited to establish virtual intelligent spaces where all outside and inside effects are reacted by a high level model. The objective is simulation of physical intelligent spaces. The authors conceptualized virtual intelligent space as a well-organized structure of descriptions and behavior definitions for engineering objects that identified in a physical space. Sensor signals in physical intelligent space are replaced by received change information, processing of change information is done by behavior analysis, and actuators are replaced by adaptive action generators. This paper starts with characteristics, of virtual intelligent space. Following this, description of behaviors of modeled objects as essential technique of virtual space is explained. Finally, essential processes in virtual space are outlined.