Heterogeneous Telescope Networks

Frederic Hessman

Transcript

1. Heterogeneous
   Telescope Networks
   Frederic “Rick” Hessman
   Georg-August-Universität Göttingen
   

   View Slide

2. The (original) Background
   •  Time-domain astronomy demands flexibility
   •  Large telescopes are inherently inflexible
   •  Small telescopes can be easily operated more flexibly
   •  There are various scientific/institutional reasons not to
   dedicate small telescope resources to a particular
   project
   •  “All” one needs is to be able to pool resources
   January  9,  2014   LOFAR  Transient  Mee9ng   2  
   

   View Slide

3. The History
   January  9,  2014   LOFAR  Transient  Mee9ng   3  
    Exeter  2005  
    GöAngen  2006  
   

   View Slide

4. The Need
   January  9,  2014   LOFAR  Transient  Mee9ng   4  
   Gamma-­‐Ray  Bursts  
   Catalina  Sky  Survey  
   Las  Cumbres  
   LSST  
   

   View Slide

5. The Idea of a Telescope eMarket
   January  9,  2014   LOFAR  Transient  Mee9ng   5  
   HTN  Server   HTN  Client  
   Broadcast  of  poten9al  
   services  (with  details!)  
   Request  for  a  quote  on  a  
   par9cular  observa9on  
   Principle  acceptance  
   indicated    
   Order  for  a  par9cular  
   observa9on  
   Denial  or  acknowledgement  
   and  unique  ID  given  
   Status  /  Abort  request  
   Status  info  
   Report  /  Data  
   

   View Slide

6. The Protocol : Remote Telescope Markup Language
   January  9,  2014   LOFAR  Transient  Mee9ng   6  
   abort        offer  
   acknowledged    reject  
   complete        report  
   confirm        request  
   fail          resource  
   incomplete      update  
   inquiry  
   e.g.  White  &  Allan  2008  
   

   View Slide

7. The Reality / The Lessons
   •  It takes a while (longer than one thinks) to get any
   system running, much less a heterogeneous one
   •  Career paths of major players are relevant
   •  “Simpler” networks are already very complicated
   •  API != implementation and one person’s particular
   implementation may not be another’s
   •  Good science drivers are needed to overcome the
   sociological/institutional inertia
   January  9,  2014   LOFAR  Transient  Mee9ng   7  
   

   View Slide

8. HTN’s “Dornröschenschlaf”
   •  VOEvent has matured (content & delivery)
   •  Many on-going surveys (OGLE, CSS, CRTS, PTF,
   Pan-STARRS, LOFAR, GAIA, …) need follow-ups
   •  Las Cumbres Observatory
   •  Many new “minor” players with resources and
   potential interest
   •  Need to create and exercise now the infrastructure
   needed tomorrow by even bigger surveys like
   LSST, SKA, …
   January  9,  2014   LOFAR  Transient  Mee9ng   8  
   potental  networked  science  
   thorny    problems  
   HTN  
   

   View Slide

9. The Problems
   Assume
   –  some widespread principle interest in providing/sharing
   resources without cumbersome institutional politics
   –  the software to make participation “easy” is available
   What might prevent/stop a HTN effort?
   –  science goals not well-defined enough?
   –  too much heterogeneity?
   –  too little heterogeneity?
   –  too little benefit for the effort?
   –  quality control?
   January  9,  2014   LOFAR  Transient  Mee9ng   9  
   

   View Slide

10. The Generic Proposal
    •  Create a suite of tools to help make any telescope
    (manual, remote, robotic) principally VOEvent-
    aware and HTN-compatible with a “minimum of
    effort”
    •  Encourage the creation of science-driven consortia
    to operate VOEvent-driven and HTN-connected
    networks
    January  9,  2014   LOFAR  Transient  Mee9ng   10  
    

    View Slide

11. The Specific Proposal
    •  Interested parties join forces to create an open-source HTN-
    client with
    –  an absolute minimum level of complexity (use & installation)
    –  transport exactly like VOEvent
    –  example of manual and automatic interfaces practically “out
    of the box”
    –  some help in interfacing (e.g. simple mapping of schemata)
    –  configurable for multiple consortia
    –  define minimum standards of client calibration/participation
    •  Create an open-source HTN-Server to foster creation of
    consortia
    January  9,  2014   LOFAR  Transient  Mee9ng   11  
    

    View Slide

12. The Summary
    •  The original HTN idea was slightly ahead of its time
    •  Things have changed/progressed
    •  We need to respond now to the “follow-up crisis”
    •  The sociological/institutional compunctions against
    joining a heterogeneous network can be overcome
    •  We just need to make it easy to join one or more
    networks by providing standard tools and simple
    implementations
    January  9,  2014   LOFAR  Transient  Mee9ng   12  
    

    View Slide