How to mobilize the agile science needed in a crisis

Climate change, the COVID-19 pandemic and conflicts have all contributed to current national and global instability. In this age of polycrises, politicians and policymakers often need to make fast decisions. Scientists in academia can struggle when asked for research to support those decisions, often because their institutions are not set up to provide authoritative knowledge for rapid decision-making.

The sprint to solve coronavirus protein structures — and disarm them with drugs

One high-profile exception to this offers a blueprint for how to mobilize and do high-quality research reliably in a crisis. During the COVID-19 pandemic, researchers in academia and industry, working with policymakers, together solved the protein structures of the SARS-CoV-2 virus in record time; conducted fast, randomized controlled trials to test drugs; and quickly developed vaccines that subsequently gained rapid approval.

Now, two initiatives from UK universities revisit questions of how to do high-quality and effective collaborative research equitably, transparently and in a timely way across disciplines. Their findings need to be considered far beyond the United Kingdom. At a time of public mistrust of institutions, they show how universities could become more agile and responsive to the needs of wider society.

In early 2022, environmental scientists at the University of Oxford, UK, asked whether academic research teams could deliver the timely science needed by decision makers while working in partnership with stakeholders, such as those in government, industry, non-profit organizations and local communities. The Oxford team began testing its idea using a series of experimental grants under a research project called the Agile Initiative. The grants would last between 6 and 12 months, with funding provided by the Natural Environment Research Council, part of the country’s main public funder, UK Research and Innovation (UKRI).

The first experiment awarded grants through a business-as-usual approach, with proposals for research topics coming from scientists. A second flipped the model, with decision makers posing questions for researchers to answer. The third was a combination, incorporating what is known as co-creation or co-production: decision makers and researchers collaborated in an iterative process on study design and implementation. The first two approaches resulted in relatively few effective ideas. The third generated strong interest from both groups and produced many good ideas, the Oxford team found (see go.nature.com/4y3zrbw).

In one example from 24 funded projects, researchers, government officials, fire brigades and Indigenous organizations collaborated on a handbook about how to fire-proof the Amazon forest. This came about after the region’s severe fire season of 2023–24, in which a drought fuelled some of the worst fires in two decades (C. Bourgoin et al. Biogeosciences 22, 5247–5256; 2025). The handbook was then distributed by boat along major rivers to hundreds of local fire-fighting teams before the next fire season, which is expected to be fierce because of the current El Niño.

Institutional barriers

The Agile Initiative also highlighted institutional barriers to research with partners outside academia. This was the subject of a separate, two-year investigation called the Co-Production Futures Inquiry. In response to a call for evidence on the scale of the problem and possible solutions, a team led by researchers at the universities of Liverpool, Manchester and Sheffield reviewed around 100 submissions from more than 90 organizations. Funded by Research England (part of UKRI), the inquiry has now released its findings in a report (see go.nature.com/46tzjep) and discussed them in a podcast (see go.nature.com/4hgdswn).

Partners outside academia need to be paid, but a common complaint among the respondents was the difficulty of making this happen. Many universities are not set up to fund individuals who are not in an academic post or employed in an established organization. Often, project teams find workarounds to pay research partners — such as temporarily putting recipients on payroll systems or searching for organizations to employ them while working with a university.

The lightning-fast quest for COVID vaccines — and what it means for other diseases

A different problem — common to both initiatives — was how to align co-created science with existing academic career structures. Early-career researchers working on agile projects were typically on short-term contracts, including those leading project teams. But a 6- or 12-month grant is precarious, even by the standards of early-career research.

Most funders require results to be published as peer- reviewed studies as a condition of awarding grants. Both initiatives found that this is not always possible, especially on a fast timescale. Publication of a peer-reviewed study can take longer than 12 months, and researchers who don’t work in universities don’t typically publish their findings in journals. For them, an effective output can be a speech, book or even a new collaboration. Co-creation and co-production also involve sharing credit across a team, something that doesn’t sit well with PhD programmes, for example, which are structured to reward individual achievement.

One of the most crucial lessons was the importance of being allowed to experiment: the idea that it is OK to fail, and to fail often before you succeed. This goes against a lot of conventional grant-making, which expects applicants to demonstrate a linear path to success. Importantly, science funders and university leadership teams were closely involved in both of these projects. This should mean that recommendations from the two reports will be taken seriously. They need to be if the world is to give itself the best chance of responding successfully to unexpected crises — regardless of their location and form.