Building national laboratories to meet China's development challenges

China boasts one of the largest scientific forces in the world, but most research institutes focus on a specialized subject of research-which many say are insufficient to meet the country's complex development needs. As part of the reform of scientific institutions and the implementation of innovation-driven development strategies, the Chinese government plans to build several comprehensive national laboratories that it hopes will further boost its scientific research prowess and to support large-scale projects.


Joël Mesot Director of Paul Scherrer Institute in Villigen, Switzerland
Tieniu Tan (Chair) Vice President of Chinese Academy of Sciences, China

THE MAKING OF NATIONAL LABORATORIES
Tan: China already has a few national laboratories. How are they compared with those, say, in the US? Ding: I spent seven years at the Department of Energy (DOE)'s Argonne National Laboratory in the US and now heads one of the national laboratories set up by the Ministry of Science and Technology of China. Most of the existing national laboratories in China are single-purpose, focusing on a particular area of research, whereas US national laboratories tend to be multi-purpose, involving research across many disciplines. There are also differences in scale. Our national laboratory, for instance, has only 500 people, whereas Argonne has over 4,000. Tan: So China needs to move on to a new generation of national laboratories that are more ambitious in both scale and undertaking. Ding: Indeed. It is very important to build comprehensive national laboratories. Compared to laboratories of a single discipline, comprehensive national laboratories can undertake more large-scale scientific projects and interdisciplinary research in areas that are driven by urgent national needs. Tan: What are the important elements of such national laboratories? 388 Natl Sci Rev, 2016, Vol. 3, No. 3 FORUM Comprehensive national laboratories can undertake more large-scale scientific projects and interdisciplinary research in areas that are driven by urgent national needs.
-Hong Ding Gibbs: In broad terms, national laboratories use science and technology to address key issues of national interest, such as economic security, national security and environmental security, consistent with the DOE's mission. National Labs have science and technology facilities that few universities or industrial organizations could afford, and do research all the way from basic research to applications. Although there is a system of national laboratories addressing the DOE's overall mission, each laboratory has its own vision and mission depending on its capabilities. Ding: National laboratories should focus on ambitious, longterm research, which cannot be done by universities or enterprises. Masiero: While national labs are mission-driven, they must be flexible enough so they can evolve with time. It's important to partner with local universities and local industry, to create a synergy between different national laboratories, and to position regional hubs in the context of national needs. Thomas Mason (Director of DOE's Oak Ridge National Laboratory in Oak Ridge, Tennessee): Even though national laboratories are national in the sense that the programmes they execute are focused on national missions, the facilities can be national or international in terms of their use. In addition, most national laboratories have a regional character in terms of their partnerships-whether it's with universities or industry or local governments. This regional character is particularly important for big countries such as China and the US, where different parts of the country are quite distinct in terms of their interest and needs and industrial characteristics. Yifang Wang (Director of Chinese Academy of Sciences' Institute of High Energy Physics in Beijing): I think there are different kinds of national labs. It's very hard to exactly copy one model or another because every country is different. Our institute was set up over 40 years ago as a single-purpose laboratory on particle physics. In the past 15 years, we've been diversifying our research focuses to become a multi-purpose national laboratory-though it's not officially called as suchnot only in terms of basic science but also applications. We are working on astrophysics and five satellites; we are building the world's largest underground laboratory, planning a circular collider, and a cutting-edge light source that will be similar to the Advanced Photon Source at Argonne; we also strive to extend our in-house research to areas such as biology, nanoscience, and environmental science. Mesot: Building a large-scale facility takes many years after the initial financial commitment. Then there is inflation and the risk of running over the budget. This has to be taken into account-and long-term, stable budget is important. You are making a commitment over a long period of time, and you'd be measured at your very first success. If the success is not delivered soon enough, there would be all sorts of implications. This has to be very well thought out. Masiero: National laboratories concentrate both financial and human resources. This is unique. In Italy, universities don't have the same level of fund, but they are free to use the facilities at national laboratories. National laboratories have to have a research priority, and provide a continuum over time-with a constant flow of resources. They serve the scientific community, and are places where many graduate students go for part of their thesis. Mesot: There is also a delicate balance between national laboratories and international organisations. There is a tendency to tone down activities in national laboratories and focus everything on international organisations-CERN (the European Organisation for Nuclear Research) is an example-but they should also complement each other. National laboratories have a very important role as a provider of facilities, equipments and staff. Meanwhile, they should make the best use of international organisations, turning them into a real advantage.

RELATIONSHIP BETWEEN NATIONAL LABORATORIES AND UNIVERSITIES
Tan: What should be the relationship between national laboratories and universities? Mesot: We should be careful that national laboratories do not duplicate university research. Rather, they should be complementary to universities. This is crucial. We should establish partnerships with universities. In Switzerland, it's quite easy because universities and national laboratories are part of the same organisation. For Chinese Academy of Sciences (CAS), it's very important to integrate your existing institutes in the future setup of national laboratories.

Hong He (Beijing representative of the Helmholtz Association of German Research Centres):
In Germany, Helmholtz is unique. We are driven by national missions, operate big science infrastructures, and try to find solutions for big challenges. As national research centres, we have big advantages in funding compared to universities that are mainly funded by local governments. And we have a stably increasing budget as institutional funding-90% of which comes from the federal government. We have to face attacks from universities and institutions such as Max Planck Institutes, which complain that we do not produce as many publications as universities for the same amount of investment, and that we do not involve enough young scientists. We are, therefore, not allowed to apply for grants from the German Science Foundation (DFG) unless the proposed project is led by a partner group from a university. Mesot: This is a fundamental issue. We were in this situation before in Switzerland, but are now allowed to apply for national science foundation grants and all kinds of European grants. This was a key turning point for us. Tan: What's the percentage of regular budgetary money in Paul Scherrer Institute? Mesot: It's about 70%. But it's not just about money. It's also about the competitive process. The fact that our scientists are allowed to apply for national science foundation grants gives them the same opportunities as university researchers, and they started to become known because the proposals are read by the community. So we don't treat scientists from national laboratories differently. And we need to make it clear that universities have free access to the facilities of national laboratories. Victoria Wright (Director of Science and Technology Facilities Council, Research Councils, UK): There are similar tensions in the UK between universities and those that run large-scale facilities with big budgets. The way we address this is to have a committee representatives of all research councils in the UK who appreciate the importance and the need to run large-scale facilities. They jointly decide how much of the total planned budget will be spent on those facilities in ways that will meet everyone's need. Ding: After spending 7 years in the Argonne National Laboratory and 10 years in US universities, I have some insights into both. In Argonne, we have facilities but no graduate students, whereas in universities, we have students but need equipment. I think CAS is unique in the sense that it has both. It's a great advantage. He: The Helmholtz Association has 18 national labs. Some of them are about 60 years old, whereas others are much younger. The establishment of the association is a direct result of a series ongoing reforms aiming to manage those big scientific facilities for restricted government funding for both scientific excellence and strategic relevance. We have to promote internal competence as well as internal and external collaboration. It's a big challenge to take care of different interests of the society, the government, and individual scientists. I think China faces similar challenges. So if you want to propose big national laboratories to the leadership in China, or if you want to win support from Chinese universities, you have to prepare arguments for that.

GREAT EXPECTATION: COMBINING BASIC AND APPLIED RESEARCH
Tan: How should national laboratories strike a good balance between basic and applied research? Wang: I think national laboratories focusing only on basic research may have troubles in the future because there are good times and bad times for science in general, especially particle physics. It's good to have long-term applied research, like light source development we are undertaking, alongside basic research. I don't think it's a good idea to separate them. Having a combination of the two, and having researchers together in the same campus, talking to each other, working together, it's beneficial. I hope CAS will have this in mind when conceiving the future of national laboratories.

Jinghua Cao (Director of CAS Bureau of International Collaboration in Beijing):
China is in the process of using alternative energy to address its complex development needs. It hopes that, by creating national labs, we will be able to address some of the challenges in the long term.
Most national laboratories [in the US] have a regional character in terms of their partnerships-whether it's with universities or industry or local governments.

-Thomas Mason
Gibbs: Many national laboratories in the US are interested in doing work that generates applications, leading to jobs and economic benefits, in addition to their other science and technology goals. I have learned at this meeting that Chinese researchers also share this goal. Mesot: Yes, there are very big expectations. There is a big risk that the expectations are on a very short time scale. We increasingly feel this pressure. The innovation chain in industry is completely different from that in basic research. We are thinking typically of 10-15 years. They are thinking of 2-3 years. To match this expectation is challenging. Basic research is key. And it takes a long time to create innovative industrial products from basic research. Tan: It's a very important point. There should be a healthy balance between basic science and applied research. We also need to have mechanisms of long-term funding to support long-term missions. Most funding cycles are 3-5 years. It's really challenging to address the balance between the two. Peter Littlewood (Director of DOE's Argonne National Laboratory in Lemont, Illinois): I don't see a real distinction between basic and applied research. The way to have an impact on big national problems is to take problems of the scale and importance which have to be solved over decades, and which give rise to the long-term vision, and approach that using fundamental science, and gradually get there. We all know that we can get science out of technology as often as we get technology from science. A good example is the discovery of a strange phenomenon called 2D electron gas, which was thought to be not terribly interesting. It was discovered when researchers at the Bell Laboratories and IBM were trying to grow perfect crystals in order to make faster transistors. Once you have perfect crystals, you could make perfect 2D electron gas, and you could not actually fail to discover some fundamental new physics.

MANAGING NATIONAL LABORATORIES
Tan: What's the best way to manage national laboratories? Mason: An important characteristic of national laboratories is the collection of people and resources together at one location. I think location is more important than organisational structure. For CAS, a big challenge is that you already have labs, institutes and facilities-the building blocks of national laboratories. One of the things that is missing is co-location, so researchers can work closely together. You can do it at a distance, but it's much harder. We move people to the same location whenever FORUM possible. Climate change scientists, computational scientists and environmental scientists, for instance, don't normally talk to one another. But if they are in the same building, they have lunch together, they argue and have discussions, pretty soon they work together in a way they wouldn't if they are several kilometres apart. Ding: Indeed. CAS has over 100 institutes on various areas of research, which should create some regional research base between institutes, universities and industry. We are building national laboratories that may be based on different models. Having regional national laboratories is in theory a good model to pursue, but how to combine existing institutions and future plans is challenging.
It's important to emphasize that national laboratories must have an open access policy-that is, their facilities have to made available to anyone whose project proposes are selected by the international scientific committees of the laboratories in question.
-Antonia Masiero Wang: I don't think distance is always an issue. It really depends on whether people share the same vision and can work together as a team. We've built large facilities in Dongguan and Kaiping, both in Guangdong province, 3,000 kilometres away from Beijing. It's not difficult with today's communication technology. We can manage this kind of projects over distance. If they can't work together as a team, it doesn't help even if they are in the same office. A lot of particle-physics project are good examples. They always have large facilities scattered around the world. Then we put them together. It's the same experiment. Distance is not a fundamental issue, not for us at least. Littlewood: I want to share with you our challenges of running a national programme called the Joint Centre for Energy Storage Research, which involves a lot of individual researchers, universities, national laboratories, and industry. We tried to drive a model of research that is quite different from the traditional approach. It's really hard. We started this by emails, telephone calls, voice mails, and very fancy video systems. But we had to give up on that. We basically had lots of face-to-face team meetingssome weekly and others monthly-and people flew around all the time. If you want to change directions, then you need a lot of discussions. And the amount of work that you need to do socially to get people to change their way of doing things is enormous. So, if you take a project and really want to push hard, then you would need very tight cooperation. This is hugely challenging. We spent a lot of time managing from the top. We spent a lot of time thinking about what is the model by which we run research. Sometimes you can really have a CERN (the European Organisation for Nuclear Research)-like collaboration that is run with military precision, so everybody knows what they have to deliver at a particular time. Every project is different.

FUTURE OF NATIONAL LABORATORIES IN EUROPE AND THE USA
Tan: What is the future of national labs in Europe and the US? Masiero: Each European country has its own research agencies and national laboratories-though their scale is smaller or much smaller than those in China or in the US. There is now an effort by the European research agencies to push forward pan-European infrastructures. The most remarkable example of such efforts is CERN, which runs the world's largest high-energy accelerator.
Upcoming facilities include the European Spallation Source [a state-of-the-art super-microscope]. The relevance of such large pan-European infrastructures is a key aspect of the present 7 th Framework Programme [a key EU funding programme] for research and technology in Europe. Indeed, in new programmes like Horizon 2020, a lot of emphases are put on innovation, on inventing new technologies, and on the scientific and technological transfer between research institutions and industry by concentrating human and financial resources into relevant, largescale European infrastructures. Littlewood: In the US, we are having an interesting discussion about the future of neuroscience, which is currently dominated by short-term medical concerns. There is an opportunity now to do imaging on a very large scale to come up with a blueprint of neural connectivity, which can then enable huge growth of all sorts of science. A historical example is the enabling technology produced by DOE national laboratories to sequence the human genome. They drove the programme at a time when medical and biological communities thought it wasn't of much interest and that it was competing with them for funds, and DOE had to fight very hard to get this on the agenda.
So I see the goal of national laboratories as being that they should be able to nucleate long-term projects that they understand they have to drive through, they have to be persistent, they may have to fight, and they may find communities against them that are established and do not want those things to happen. Only in this way can national laboratories catalyse sizeable and important progress.

Xueming Yang (CAS Dalian Institute of Chemical Physics):
We're in the process of establishing of a national laboratory for clean energy. The challenge is to identify national needs and national strategies. In the US, it seems that multi-disciplinary laboratories are designed to serve national needs. What's the assessment by the US government in terms of how the nationallaboratory system is working? Mason: This is a very topical issue. There have been many studies in the past few years. Most of the areas with the most acute problems that stimulate some of the studies are in the nuclearsecurity space, and some of those laboratories have been the focus of a lot of controversies and criticisms. For other labs, the problems are not so acute, and it's considered that the planning and the stewardship generally work well.
One of the biggest concerns is cost-effectiveness, particularly given the budget pressure in the past few years. Some would say that maybe there are too many national labs. But, if you read the