Automated research workflows (ARWs) could vastly speed the pace of discovery and hence science’s contributions to society. A National Academies of Sciences, Engineering, and Medicine report calls for research institutions, funders, and other stakeholders to invest more in and share these methods with the wider scientific community.
ARWs combine computation, laboratory automation, and artificial intelligence tools to streamline experiments and data analysis. They have already accelerated research in a wide array of disciplines. For example, materials science researchers are using ARWs to cut the time needed to synthesize and test materials from 9 months to 5 days, according to the report. And in drug discovery, an algorithm has increased the percentage of identified active compounds in a data set to 57%, up from 20% using traditional models.
The report recommends that federal funders encourage the creation of tools, platforms, and data archives needed to support ARWs while at the same time making sure the resulting research is done ethically and openly. Funders, higher education institutions, and scientific societies also need to design educational programs to create the workforce needed to develop and use ARWs.
The main challenges of ARWs include the “long-term sustainability of cyberinfrastructure (computing, networking, and now critically, data and programs); reducing barriers and increasing incentives for interdisciplinary collaboration; addressing security challenges; and educating current practitioners and students to design and responsibly use ARWs,” University of Michigan emeritus professor Daniel E. Atkins, chair of the committee that produced the report, writes in the preface.