Infrastructure systems and smart buildings are rapidly joining the Internet of Things and evolving into advanced cyber-physical systems. As a result, massive amounts of data that characterize the structure and function of urban areas in minute detail are being generated. However, these data are often fragmented and managed by a variety of public agencies and private corporations. As a result, they are not readily available to the urban research community. This paper lays out a strategy to develop a data commons that would collect, curate and distribute Urban Big Data to support research on infrastructure systems and how they interact with the human populations they support.