Share Email Print
cover

Proceedings Paper

Towards a dynamical scheduler for ALMA: a science - software collaboration
Author(s): Jorge Avarias; Ignacio Toledo; Daniel Espada; John Hibbard; Lars-Ake Nyman; Rafael Hiriart
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

State-of-the art astronomical facilities are costly to build and operate, hence it is essential that these facilities must be operated as much efficiently as possible, trying to maximize the scientific output and at the same time minimizing overhead times. Over the latest decades the scheduling problem has drawn attention of research because new facilities have been demonstrated that is unfeasible to try to schedule observations manually, due the complexity to satisfy the astronomical and instrumental constraints and the number of scientific proposals to be reviewed and evaluated in near real-time. In addition, the dynamic nature of some constraints make this problem even more difficult. The Atacama Large Millimeter/submillimeter Array (ALMA) is a major collaboration effort between European (ESO), North American (NRAO) and East Asian countries (NAOJ), under operations on the Chilean Chajnantor plateau, at 5.000 meters of altitude. During normal operations at least two independent arrays are available, aiming to achieve different types of science. Since ALMA does not observe in the visible spectrum, observations are not limited to night time only, thus a 24/7 operation with little downtime as possible is expected when full operations state will have been reached. However, during preliminary operations (early-science) ALMA has been operated on tied schedules using around half of the whole day-time to conduct scientific observations. The purpose of this paper is to explain how the observation scheduling and its optimization is done within ALMA, giving details about the problem complexity, its similarities and differences with traditional scheduling problems found in the literature. The paper delves into the current recommendation system implementation and the difficulties found during the road to its deployment in production.

Paper Details

Date Published: 26 July 2016
PDF: 14 pages
Proc. SPIE 9913, Software and Cyberinfrastructure for Astronomy IV, 99133W (26 July 2016); doi: 10.1117/12.2232004
Show Author Affiliations
Jorge Avarias, Joint ALMA Observatory (Chile)
National Radio Astronomy Observatory (United States)
Ignacio Toledo, Joint ALMA Observatory (Chile)
Daniel Espada, National Radio Astronomy Observatory (United States)
John Hibbard, National Radio Astronomy Observatory (United States)
Lars-Ake Nyman, Joint ALMA Observatory (Chile)
Rafael Hiriart, National Radio Astronomy Observatory (United States)


Published in SPIE Proceedings Vol. 9913:
Software and Cyberinfrastructure for Astronomy IV
Gianluca Chiozzi; Juan C. Guzman, Editor(s)

© SPIE. Terms of Use
Back to Top