Dynamic scheduling for SOXS instrument: environment, algorithms and development

TL;DR

This paper discusses the development of an automated, fast-response scheduler for the SOXS instrument at ESO-NTT, designed to optimize target observation plans under changing conditions for transient astrophysics.

Contribution

It introduces a new dynamic scheduling algorithm tailored for the SOXS instrument, addressing the challenge of automated, real-time observation planning without on-site astronomers.

Findings

Scheduler is nearly complete and functional.

Web interface for scheduler is developed.

Scheduler provides optimal target and backup lists.

Abstract

We present development progress of the scheduler for the Son Of X-Shooter (SOXS) instrument at the ESO-NTT 3.58 meter telescope. SOXS will be a single object spectroscopic facility, consisting of a two-arms high-efficiency spectrograph covering the spectral range 350-2000 nanometer with a mean resolving power R$\approx$4500. SOXS will be uniquely dedicated to the UV-visible and near infrared follow up of astrophysical transients, with a very wide pool of targets available from the streaming services of wide-field telescopes, current and future. This instrument will serve a variety of scientific scopes in the astrophysical community, with each scope eliciting its specific requirements for observation planning, that the observing scheduler has to meet. Due to directions from the European Southern Observatory (ESO), the instrument will be operated only by La Silla staff, with no astronomer present on the mountain. This implies a new challenge for the scheduling process, requiring a fully automated algorithm that should be able to present the operator not only with and ordered list of optimal targets, but also with optimal back-ups, should anything in the observing conditions change. This imposes a fast-response capability to the scheduler, without compromising the optimization process, that ensures good quality of the observations. In this paper we present the current state of the scheduler, that is now almost complete, and of its web interface.