Astro

Consolidated R&D&I project


ASTRO-IA

Spectroscopic Analysis and Recognition Techniques Optimised with Artificial Intelligence (ASTRO-IA)


Reference
CIAICO/2024/186
Funding body
Generalitat Valenciana · Regional Ministry of Education, Culture, Universities and Employment
Call
Grants for consolidated research groups AICO 2025 (CIAICO 2024)
Period
2025-2028
Principal investigators
Ignacio Negueruela Díez and Amparo Marco Tobarra

The project

The arrival of large spectroscopic surveys represents a revolution in astrophysical techniques, bringing new demands and procedures. Analysing hundreds of thousands of spectra requires a level of automation beyond the capabilities of current codes. Although analysis tools are being developed for stars considered "normal" (Sun-like dwarfs and red giants), until now massive stars have lacked dedicated tools, despite their crucial role in understanding the physical processes that shape our Universe.

The Stellar Astrophysics Group of the University of Alicante has developed Astro+, an advanced application for analysing the spectra of massive stars, both hot and cool. At present, Astro+ can handle samples of many hundreds or even a few thousand spectra. To tackle the scale of projects such as the WEAVE/SCIP survey, which will provide tens of thousands of spectra, new artificial intelligence tools must be integrated into Astro+ to speed up every process and bring computing times down to acceptable levels.

The aim of this project is to develop and apply advanced machine learning techniques to accelerate and scale up spectroscopic analysis in Astro+. To this end, in collaboration with experts in advanced machine learning, we are carrying out a thorough, in-depth study of methodologies for one-dimensional signal processing, adapting them to spectral analysis. The work follows a sequential plan focused on developing and training models able to identify spectral properties efficiently, paving the way towards even more sophisticated tools. The project funds the hiring of a specialist technician to implement all these ideas and experiments in code, as well as other associated costs. With these advances we aim to become a world reference in the study of massive stars and automated spectroscopic analysis.

Acknowledgements

Project CIAICO/2024/186 funded by the Regional Ministry of Education, Culture, Universities and Employment of the Generalitat Valenciana.

Universidad de AlicanteGeneralitat Valenciana