The Second Life of a Sensor Integrating Real-world Experience in Virtual Worlds using Mobile Phones
Mirco Musolesi , Emiliano Miluzzo , Nicholas D. Lane , Shane B. Eisenman , Tanzeem Choudhury and Andrew T. Campbell
Dartmouth College, USA Columbia University, USA
physical attributes. Avatars may be completely invented, without any relation to the physical aspect of the human user, or can be made to resemble the human whom they represent. However, in the current system, there is still a divide between the virtual world and the real one. Specifically, a user may be able to recreate his appearance but he can not automatically mirror/translate his current actions and movements in his real life to the virtual life of his avatar. Our vision is to bridge the current divide between real worlds and virtual worlds, by having real world activities (e.g., sitting, walking, running, cooking) and other aspects of the human represented in real time in the virtual world. Our aim is to go a step further than simply reproducing spaces (e.g., rooms, buildings, public spaces [7, 11]) and the objects with which humans interact. We focus instead on the provision of the virtual representation of humans, their surroundings as sensed by the humans themselves, and interactions with the members of their social networks (e.g., fetched from external sources such as Facebook [3]). There are existing virtual worlds that map the real one with great accuracy in terms of geographical locations and local features. Complex virtual representations of real geographical places exist such as the ones provided by Google Earth [4] and Microsoft Virtual Earth [9]. In Google Earth it is possible to insert 3-D representations of buildings and artifacts by means of Google Sketch. We believe that the next step will be representing real world activities in this kind of virtual setting whereby personal avatars will be able to move inside these articial spaces following the real positions and activities of people. Sensors have been used to support presence in virtual environments since the rst experiments on the Virtual Reality Theaters [2]. In most of these systems, individuals carry special wearable devices like helmets and wristbands with embedded sensors measuring their movements. These are special-purpose external devices used to map user actions. More recently, the integration of sensors and Second Life is discussed as a future work in [7], where the authors propose to reproduce the entire oor of a building using data collected by means of the Plug [7] platform. However, the focus of [7] is on the reproduction of a working space and not on the activities of people moving inside it (i.e, it is focused on places and objects and not on subjects). Moreover, [7] is based on a non-standard hardware platform produced in small quantities and not available to the general public. Starting from this technological scenario, the contribution of this paper can be summarized as follows:

下一页