An Anglo-South African project is employing AI and digital technologies to automate optical and infrared telescopes and make them “smarter”, that is, able to monitor themselves, immediately process data, and capture images of sudden astronomical occurrences, such as exploding stars, in real time. The programme, designated “Intelligent Observatory”, is a collaboration between the Hartree Centre of the Science and Technology Facilities Council of the UK, and the South African Astronomical Observatory of the National Research Foundation (NRF-SAAO). It is being funded by the UK Research and Innovation International Science Partnership Fund.   

“It’s incredibly exciting to work with the [SAAO] on the Intelligent Observatory programme, showing how advanced AI can transform scientific discovery, making telescopes smarter, enabling astronomers to focus on exploration and breakthrough discoveries,” highlighted Hartree Centre principal AI researcher Dr Adriano Agnello. “These tools mark a significant step forward in the application of AI to observatory operations, broadening access to major scientific facilities and supporting translational research that can also be applied to industrial settings.”

“By combining the Hartree Centre’s world-leading expertise in AI and high-performance computing with NRF-SAAO’s deep experience in telescope operations, the Intelligent Observatory represents a fundamental shift in how astronomical facilities operate,” stressed NRF-SAAO astronomy head Professor Stephen Potter. “Together, we are embedding advanced AI into the heart of our telescopes and data systems, building observatories that are not just automated, but truly intelligent. This partnership strengthens South Africa’s leadership in innovative astronomy while creating powerful opportunities for students and researchers across Africa to engage with cutting-edge AI-driven science.”

The combined team is developing software to monitor telescopes and spot and flag any problems early, as well as to convert raw observation data into “clear insights”, with the minimum of human input. It will summarise nightly observations and flag weather disruptions, and so make certain that the astronomers collect the best data possible. The advanced AI models will forecast local weather, make adjustments to cope with atmospheric effects, and detect faults, thereby ensuring smooth telescope operations and reliable observations.

The team has already built, using Large Language Models, a smart text search platform, to search and interpret technical documents. It is hosted at the Hartree Centre, included simple feedback interfaces, managed AI issues and reduced incorrect responses. It gives astronomers fast and accurate answers even during busy observing nights. The next phase will be the addition of Southern African Large Telescope documents, and the application of advanced techniques to increase the intelligence and power of the platform.   

“Beyond astronomy, observatories are perfect testbeds for new technologies that can help industries such as manufacturing, energy and transport, from smart sensor technologies to predictive maintenance and automated data systems,” pointed out the NRF. “By lowering barriers to information, these tools also give students, including those at historically disadvantaged institutions, important hands-on experience in AI and data science.”