Institutes & Centers
Thanks to an explosion of astronomical data and the rapid development of computing technology, a paradigm shift in scientific research is taking place in the field of astronomy. On November 6-7, 2023, the International Workshop on Intelligent Computing in Astronomy was held at Zhejiang Lab (ZJ Lab). More than 300 "stargazers" from five countries, including Australia, France, Germany, Japan and China, gathered in Hangzhou to discuss how to achieve better integration of intelligent computing and astronomy, so as to promote scientific discoveries and technological innovations.
"ZJ Lab is always committed to Computing for Innovation. We hope that astronomers and computational scientists can work together to explore how computing contributes to scientific research and enables computing-driven scientific discoveries and technological innovations," said WANG Jian, President of ZJ Lab and Member of the Chinese Academy of Engineering, in his remarks.
Computing, an indispensable technological force for astronomical research
"As we enter a new era of space exploration, one of the toughest challenges we face is processing mass data generated by observations. The adoption of advanced data science and technology and intelligent computing is considered not just as an auxiliary role, but the key to unlocking new theories of cosmology," said HUANG Song, Assistant Professor of the Department of Astronomy at Tsinghua University. LIU Chao, Researcher of the National Astronomical Observatories of the Chinese Academy of Sciences, pointed out that AI algorithms are widely used in many areas of astronomy, and it has great advantages when a large number of observations and analytical calculations are processed.
At the workshop, domestic and foreign experts delivered wonderful talks about astronomical computing. Matthew Bailes, winner of the Shaw Prize in Astronomy and Director of the Center of Excellence for Gravitational Wave Discovery of the Australian Research Council (ARC), described how pulsars and fast radio bursts were discovered and relevant data were processed. LIU Chao took the Chinese Space Station Telescope (CSST) as an example to demonstrate how AI algorithms assist astronomers with large-scale studies containing tens of millions of images and about 10 billion objects. HUANG Song introduced the MUltiplexed Survey Telescope (MUST) program, and dove into its pivotal role in the upcoming fifth phase of cosmic surveys.
Takahiro S Yamamoto, a postdoc in the cosmology group at Nagoya University who is actively exploring the application of deep learning in gravitational-wave astronomy, reviewed in his talk how AI was used to search for continuous gravitational waves. Valentine Wakelam, Deputy Director of Laboratoire d'Astrophysicals de Bordeaux, presented new-generation complex chemical models for interpreting interstellar medium observations, and showed how to develop methods for processing, visualizing and interpreting these observations. WANG Lile, Assistant Professor at the Kavli Institute for Astronomy and Astrophysics at Peking University, presented some key problems in computational astrophysics and how to solve them, and shared the research progress in heterogeneous computing and "AI for science".
Ting Yuan-Sen, Associate Professor of Astronomy and Computer Science Departments at Australian National University, presented AstroLLaMa, a large language model (LLM) in astronomy. He pointed out that LLMs can generate scientific hypotheses comparable in complexity to human-generated hypotheses through context-sensitive help and fine-tuning on domain-specific documents, and that these specialized foundation models can revolutionize the methods we use to retrieve documents and track domain-specific knowledge. Prof. GAO Yiqin at the College of Chemistry and Molecular Engineering of Peking University demonstrated the replication and training from scratch (or de-novo training) of protein structure AI models based on the software and hardware framework fully developed by China, and presented AI-native multimodal and cross-scale molecular modeling tools. HUANG Boyuan, a Senior Product Expert of Alibaba Cloud Intelligence, presented the AI machine learning platform that can be applied to various industry scenarios. LIU Yiming, a Senior Product Expert of Alibaba Cloud Intelligence, presented the Open Data Processing Service (ODPS) developed by Alibaba Cloud, and demonstrated its convenient, efficient and scalable data processing capabilities based on typical use cases of biomedicine, astronomy and meteorology. CHEN Hongyang, Deputy Director of ZJ Lab's Research Center for Graph Computing, presented the development process of graph computing technology and how to empower scientific discoveries with this technology, and showed the recent work on the Zhuque Graph Computing Platform and the Zhuque Graph Pretrained Foundation Model.
At the roundtable session hosted by LI Di, Chief Scientist of FAST and Chief Scientist of Computational Astronomy at ZJ Lab, keynote speakers had an open discussion around the theme "Computing Senses Cosmos".
We can see further only when more people get involved
During the two-day workshop, the speakers in the field of astronomy shared their achievements in promoting innovation in the integration of astronomy and computing in their respective fields, and the speakers in the field of computing fully demonstrated the potential of computing to support astronomical research. The sharing and communication extended from the meeting venue to corridors and luncheons, which could be found everywhere.
As Academician WANG Jian said in his remarks, "Getting inspiration from face-to-face communication is more important than reading papers. I hope that more interesting discussions will happen through such workshops (I prefer to call it a workshop rather than a conference). ZJ Lab is creating an open science environment where good digital infrastructure and software and hardware environment will be built up to support scientific research, hoping to engage scientists in more fruitful discussions."
"As an astronomical researcher, my routine work is to search for valuable information from massive astronomical data. So, how to work faster and more efficiently is a problem we have been trying to solve. This offline discussion gives astronomical researchers a good opportunity to dig into and solve problems with computational researchers." YAO Jumei, an Associate Researcher at the Xinjiang Astronomical Observatory of the Chinese Academy of Sciences, shared her feelings about this workshop.
To look up at the stars and look into the future, we need to go the extra mile and exert more effort on scientific research, science communication, and science popularization. At the closing ceremony of the International Workshop on Intelligent Computing in Astronomy, ZJ Lab Development Foundation formed a partnership with the National Astronomical Data Center. The two sides will jointly promote opening sharing of scientific and technological resources, spread scientific knowledge across the society, carry forward the spirit of science, and light up science and technology for the common good. The Center for Publicity and Culture of ZJ Lab cooperated with OPENVERSE to build scientific arts studio "Intouchable in Zhejiang Lab", so as to jointly empower science communication through science-arts integration.
YUAN Jixin, Vice President of ZJ Lab, presents "Public Welfare Partner of ZJ Lab Development Foundation" plaque to CUI Chenzhou, Director of the National Astronomical Data Center
DENG Mingyue, co-founder of OPENVERSE, introduces the development vision of Intouchable in Zhejiang Lab
In his closing remarks, LI Di said that due to uncertainties in astronomical research itself, it is difficult to predict what changes will take place in the next decade. However, we firmly believe that, with the empowerment of computing and the collaboration of scientific research, we'll definitely see new progress in astronomy, or perhaps major scientific discoveries when we meet again next year.