AI quest sheds light on how to study distant planets

A global competition focused on advancing research in space science and exoplanets, has concluded with remarkable results.

The competition attracted over 1,400 participants based in 75 countries, who competed for the $50,000 prize pool. In an impressive field of competitors, six teams distinguished themselves, and will be able to present their solutions at the prestigious NeurIPS 2024 conference in Vancouver on 14 December 2024. First place was won by PhD student Kohki Horie and Master’s student Yamato Arai at the University of Tokyo.

Dr Papageorgiou said: “The worldwide interest in this year’s data challenge has been phenomenal and it has surpassed all expectations. The science of extrasolar planets and the advanced techniques that we must use to investigate them are clearly of huge interest to scientists in all disciplines. The global AI community has engaged enthusiastically with the data challenge and made it a huge success.”

Dr Mugnai added: “In science it’s always important to be open-minded and receptive to new ideas, and to take advantage of insights wherever they come from. That’s the spirit of  the Ariel Data Challenge. It helps the Ariel team to engage with experts from other fields and to benefit from the talent and know-how of data-handling experts worldwide. It is also tremendously exciting for people outside the project to learn about and contribute to a state-of-the-art space mission.”

Ariel Data Challenge Lead, Dr Gordon (Kai Hou) Yip of University College London, said: “I want to give a big shout-out to our winners who really stood out in this intense competition. I can't thank everyone enough who put in their time and energy over these past three months. Your work is what helps us break new ground in data analysis and really push the envelope of what we can achieve in this field.”

This groundbreaking challenge was made possible through a collaborative effort led by the UCL Centre for Space Exochemistry Data, bringing together an impressive international team of academic partners including the Centre National D'études Spatiales (CNES), Cardiff University, Sapienza Università di Roma, and the Institut Astrophysique de Paris.

The Ariel Data Challenge 2024 was sponsored by the Kaggle Competitions Research Program and CNES. The competition also benefited from the support of a consortium of leading space agencies and institutions, including the UK Space Agency, European Space Agency, Europlanet Society, STFC RAL Space, and STFC DiRAC HPC Facility.