Virtual Reality is a subject of rising importance since it is representative of the expanding interface between human beings and computers. Kim Joung-hyun is an associate professor in computer science and engineering at Pohang University of Science and Technology (POSTECH). He has contributed the following article on virtual reality, to follow the first which appeared in the Korea IT Times April issue. In the last two articles, I introduced readers to a variety of sensors and displays used for virtual reality systems. For a given content, system developers must come up with an interactive system for users that is natural, easy to use and learn, and requires low cognitive or physiological burden (e.g. sickness). In addition, the interface must be multimodal, stimulating as many of our sensory systems as possible to enhance the user experience. Interactive design is very important in terms of deploying a successful VR system, and it is also very difficult due to the nature of the problem itself (having to satisfy multiple goals simultaneously). In addition, it is further complicated by the many possible choices for sensors and displays, constantly changing technologies, and individual differences. However, interaction design still remains to be a sort of black art without standard guidelines. Figure 1 shows a successful case of a multimodal interactive design. The figure illustrates a mixed reality-based cell phone evaluation system. The user holds a blue rectangular prop (with yellow dots on it, left of Figure 1) and manipulates it as if it were a cellphone. The position and pose of the prop is detected and tracked by a camera. Through the display, the user sees a computer-enhanced design model of the cellphone (overlaid over the prop) in his hand. The user can interact with the phone by pressing a button that is mounted on the prop, observe the phone's behavior, and appreciate its design and functionality. The user is provided with visual, aural, tactile, and even proprioceptive (movement of one's limbs) feedback. Providing multimodal feedback is only one part of providing an effective interaction; the other equally important part is the realistic simulation or the behavior of the feedback. While there are an infinite number of objects to simulate, two types of simulation are required of almost all virtual reality systems. One is the simulation of object collision, and the other is that of object motion. They are important in providing the minimum level of realism (unless the very purpose of the virtual world was to ignore physical laws). The exact detection of collision for all objects in the virtual world to their geometric details is a very computationally-demanding process. As such, most systems only simulate it at a basic level using, for instance, bounding boxes or spheres (See Figure 2). Detection of collision is also important for implementing "direct" interaction. In order for the user to touch and manipulate a virtual object directly, detection of collision between the user's hand/finger and the object must be carried out. Simulation of motion is likewise a computationally burdening process. To reduce the computational burden, object motion is often computed with simplified physical models. Motion simulation is also tied to object collision, because for a collision to occur in the first place, it must have moved, and after the collision, there would be a response motion. In fact, upon a collision, objects can even break or deform. Realistic simulation of such phenomena in real time is a hot research topic. Fortunately, according to psychologists, humans do not perceive the world like computers; in other words, it is hard for us to tell whether something, for instance, moved exactly according to Newton's law. Both appearance-wise and behavior-wise, only a minimal level of realism is often sufficient to fool the user, and humans' cognitive systems have the amazing ability to fill in the missing details. This makes many virtual reality systems effective in their application domain without achieving full life-like realism.