The National Science Foundation (NSF) has awarded Boston University $20 million to help create a national network of cloud labs, artificial intelligence-infused research facilities that can remotely run science and engineering studies. The network will allow researchers from anywhere in the country to request complex experiments and tests that they don’t have the equipment or expertise to do themselves.
The network will be made up of 20 individual labs , or nodes, around the United States, each specializing in different areas of research, from materials science to chemistry. The BU-based lab will focus on biotechnology and engineering biological systems and have the capacity to host studies exploring topics such as new vaccines, DNA sequencing, and genetic engineering.
BU is already home to a cloud lab: the Design, Automation, Manufacturing, and Processes (DAMP) Laboratory allows researchers to commission and lead studies on biological matter, chemicals, and liquids without ever stepping foot on campus. In recent years, the lab has powered through studies on a biosensor that monitors fertility, complex protein functions for therapeutics, and the relationship between plants and bacteria.
The new four-year NSF award will be used to enhance the DAMP Lab’s facilities, step up its automation capabilities, and connect it to partner labs nationwide. The NSF awarded a total of $380 million to the 20 successful teams; the nonprofit Astera Institute is also contributing upwards of $20 million.
“This project is about making biotechnology faster, more reliable, and more widely available,” says Kenneth Lutchen , BU’s vice president and associate provost for research. “It has the potential to give more researchers access to advanced lab capabilities, regardless of their location or resources. Ultimately, that means more people will have the tools they need to tackle important scientific and societal challenges in biomedicine and research.”
The DAMP Lab—which has a series of high-speed liquid handling robots at its heart—can process thousands, potentially millions, of samples with minimal human intervention. During the coronavirus pandemic, it served as the University’s primary diagnostic facility, running 6,000 COVID tests a day .
According to the DAMP Lab team, being a founding member of the NSF initiative validates BU’s vision in investing in a cloud lab—it was established in 2018. Lutchen adds that it’s also recognition of the University’s rise as a leading research institution.
“This award reflects the major investments Boston University has made to build a research enterprise capable of competing for large, complex grants,” says Lutchen. “We’ve invested in our faculty, our facilities, and our ability to bring together experts from different disciplines to tackle big societal challenges.”
The BU team’s goal is to make requesting an experiment as simple as asking a popular artificial intelligence (AI) tool to sift through data or code a software program. If a researcher asks what would happen if a particular gene were added to a bacterium, for example, the system would design a test, ensuring it meets safety guidelines, and the user could simply click a button to make it happen.
“We’re trying to lower the barrier to engagement,” says Douglas Densmore , founding director of the DAMP Lab and a BU College of Engineering professor of electrical and computer engineering. “We hope long term it’s going to democratize the space and lower the capital expenses of building up a biotechnology company.”
One part of the NSF network initiative aims to ensure the individual nodes can communicate with each other—what Densmore calls “round tripping”—enabling studies requiring different capabilities to jump from lab to lab as they progress.
“If we’re going to scale-up to tackle challenges in medicine, materials, or sensing, it can’t be that there’s a lab in California doing it one way and a lab in New York doing it another,” says Densmore, who also has an appointment in ENG’s materials science and engineering division. “We need to leverage network effects; there needs to be a standardized platform. It’s important the nation invests in this for national competitiveness.”
The BU team will include Densmore and his ENG colleagues Yannis Paschalidis and Keith Brown . Paschalidis is director of the BU Rafik B. Hariri Institute for Computing and Computational Science & Engineering and an expert on AI. In recent projects, he’s tested AI’s ability to help treat hypertension, diagnose Alzheimer’s disease, and track emerging infectious diseases. Brown, an associate professor of mechanical engineering, studies soft materials like polymers and smart fluids, but has also explored innovative ways of automating research. His lab developed an autonomous robot to search for and make the most efficient energy-absorbing shape ever.
As part of the new grant, Paschalidis will create a large language model—a type of AI program—to help process lab requests and coordinate all the steps needed for complex experiments. “My team will develop AI approaches to realize the cloud, self-driven lab vision,” says Paschalidis, “including language models to translate user descriptions of experiments into code that runs in lab equipment, lab and task scheduling automation, and mechanisms to improve AI from user feedback.”
Brown will make the lab “intelligent, self-driving,” says Densmore. “It’s a lab that can, in the absence of direct input from users, tune itself and learn what it should be doing.”
“This was a highly competitive program that required applicant institutions to have a unique combination of an established user facility, expertise in robotics for science, and AI,” says Brown, head of ENG’s materials science and engineering division. “This award shows that we have a proven track record of excellence in these areas.”
Among the proponents of cloud labs is Congressman Jake Auchincloss (D-Mass.), who says the award to BU helps put Massachusetts at the forefront of boosting American research and development (R&D). He’s a cosponsor of the bipartisan Cloud Labs to Advance Biotechnology Act, or the Cloud LAB Act , which backers say would expand access to cutting-edge research facilities, help maintain American competitiveness, and accelerate the development of new technologies in agriculture, defense, healthcare, and manufacturing.
“For decades, science has grappled with declining R&D productivity. Cloud labs could help flip that trend, and Massachusetts should be the pioneer,” says Auchincloss. “Boston University is helping ensure the best science in the world is done in the Bay State.”
By the end of the four years, Densmore hopes the team has a model that can be replicated elsewhere, a template for other DAMP Labs to fire up around the US.
“Heaven forbid we have another bio-readiness event, like a COVID,” he says. “It can’t be, ‘Let’s take a year to spin up vaccine producers across the country.’ We need to be able to bootstrap new DAMP Labs in Iowa, Nebraska, Louisiana, all using standardized interfaces, all communicating—we need to build that infrastructure.”
The project includes the following subcontractors: Ben Gordon, Asimov; Pete Kelly, Align Foundation; Kevin LeShane (ENG’14), Lattice Automation; Chris Voigt, Massachusetts Institute of Technology; Renee Wegrzyn, Transfyr.