Data captured from the physical world through receptor devices such as wireless sensor networks and RFID readers tend to be low-level, device-specific, inconsistent, and unreliable. Meanwhile, data management and query processing techniques in the digital world rely on understandable, consistent, and complete data. This paper aims to bridge this physical-digital divide by introducing Virtual devICEs (VICEs): extensible wrappers that transform the data captured from the physical world into data that can be operated upon by today's data management systems. In this sense, VICEs provide what we term metaphysical data independence by shielding the above data processing system from the complexity of real-world data. This paper introduces the specific challenges to implementing VICEs, the functionality required of them, and outlines the VICE architecture. To validate the VICE approach, we demonstrate three real-world use cases that illustrate the empirical benefits of using VICEs to provide metaphysical data independence.