Astronomy 2020: A Pragmatic Approach

In the cinema history of astronomy, we are currently at the stage of the Lumiere brothers with contemporary surveys providing short monochromatic time sequences of the sky. By the end of the next decade, however, panchromatic blockbusters will be commonplace and science will be predominantly driven by the objects that change in successive “frames”. Web-scale computing resources will be required just to process the torrents of data events but the key to understanding them will be contextualisation — linking together disparate (sets of) events and relating them to archival and supplementary data in a machine-comprehensible way. Much of the data mining and analysis of such data portfolios will be performed by proxy scientists — intelligent agent avatars that represent an individual’s particular research interests in high-dimension parameter spaces. Although this view might sound like science fiction, in this paper, I will review the technologies that will make it achievable. In particular, I will cover new approaches to web services that will be required to support these massive event streams, social networking techniques that will facilitate science and semantic technologies that will underpin everything