ADASS posters are displayed all week
When
Theme: AI in Astronomy
Telescope facilities in modern astronomical research generate a substantial volume of data, including scientific observations and a large amount of housekeeping and auxiliary information from diverse sources, such as weather stations, sensors, log messages, LiDARs (light detection and ranging), and FRAMs (photometric robotic atmospheric monitor). The multitude of sensors spread throughout these facilities makes them some Internet of Things (IoT) environments. Handling and processing this vast amount of data necessitate sophisticated software architectures, which exploit cutting-edge technologies in the field of IoT and big data. While sensor data traditionally address systematic errors in scientific measurements, this paper explores their potential for novel maintenance techniques, akin to those in Industry 4.0.
Predictive maintenance has emerged as a proactive strategy to optimize the performance and operational efficiency of complex systems. In the context of telescope arrays, where downtime can significantly impact scientific research, the application of predictive maintenance assumes critical importance. Traditional maintenance practices often rely on scheduled routines or reactive approaches, leading to potential equipment failures and costly repairs. However, by exploiting the vast amounts of data generated by telescope facilities, a new era of maintenance for telescope facilities is unfolding.
This study provides significant outcomes resulting from an in-depth analysis of historical data gathered from ASTRI Horn, a Cherenkov telescope positioned at the Astrophysical Observatory of Catania (Serra La Nave, Mount Etna). The research focuses on a comprehensive exploration of data patterns spanning seven years of telescope activities. Within this analysis, we delved into the distribution of variables and how they correlate with each other. Additionally, we varied the analysis interval's granularity to assess the correlation time scale. The findings of this analysis provide valuable insights into potential progressions in strategies for predictive maintenance within telescope facilities.