Berkeley Lab Computing Sciences Takes Novel Approach to Science Workforce Development and Diversity
Sustainable Research Pathways Program’s Impact on Young Researchers Spans Nearly a Decade
May 31, 2023
by Kathy Kincade
Contact: cscomms@lbl.gov
“The opportunity to do research at a National Lab was something I did not know was possible, especially being an international student. The icing on the cake was that, through SRP, I also got to know about different ongoing research projects in areas I had not been exposed to, like machine learning and quantum computing.” - Mercy Amankwah, Case Western University, SRP 2021
Amankwah’s story is one of hundreds that illustrate the impact the Sustainable Research Pathways (SRP) program has had on early career scientists and their mentors since its founding through a collaboration between Lawrence Berkeley National Laboratory (Berkeley Lab) and the Sustainable Horizons Institute (SHI) in 2014. SRP has been connecting students from underrepresented communities and institutions to scientists and staff at Berkeley Lab for nearly a decade to facilitate research collaborations and bolster workforce development and diversity, equity, and inclusion. Through SRP, faculty and students from predominantly minority-serving academic institutions and women’s, liberal arts, community, and other small colleges and universities have come to Berkeley Lab – and, more recently, several other national labs – each summer to experience the diverse research environment available here and work closely with world-class scientists.
“A lot of what we are trying to do is create new pathways into the labs by broadening the networks of the staff who work there and how they recruit younger researchers,” said Mary Ann Leung, founder and president of SHI. “One of our biggest accomplishments is that SRP has brought a lot of people to the labs that I think never would have gotten there without this program.”
"It is often difficult for faculty members and students from minority-serving institutions and small colleges to avail internship opportunities. The SRP program provides that vital bridge between diverse academic institutions and the government’s science workforce. The opportunity to participate in SRP/SHI has culminated in several publications, open-source software, and internship opportunities for my students. It has also enabled me to broaden my professional network, which yielded invitations for writing grant proposals, visiting scholar positions, and invited talks. The research alone has served as the basis of my DOE Early Career Research Award work." -Tanzima Islam, Texas State University, SRP 2019
In essence, SRP was borne out of a growing recognition of the need to enhance equity, diversity, and inclusion across the national labs.
“We were talking about staff diversity in the Computing Sciences Area (CSA) soon after I joined Berkeley Lab in 2011, even before,” said David Brown, a key advocate for the SRP program who retired from Berkeley Lab in 2022 after more than a decade as director of the Computational Research Division (now the AMCR Division). These conversations included Kathy Yelick, former associate lab director who led CSA from 2010 through December 2019; Deb Agarwal, director of CSA’s Scientific Data Division who has long been proactive in recruiting and building a more diverse workforce; and Silvia Crivelli, a staff scientist and program manager in AMCR’s Applied Computing for Scientific Discovery Group who, prior to the introduction of SRP, implemented her own efforts to increase workforce diversity via the Department of Energy's (DOE) Visiting Faculty Program (VFP) and the WeFold protein folding project. Crivelli has also been active in the SRP program since its inception, mentoring up to 15 students each summer, many of whom have returned to the Lab and some who are now part of her research team.
“It is so rewarding to see this evolution,” Crivelli said. “Many of the people in my group were SRP students who came and stayed, and you hear from them how much this has made an impact. There are so many success stories.”
Together, these individuals, and many others, have worked to establish and promote the SRP vision. Their efforts have included addressing processes and attitudes around recruitment, hiring, and retention across the DOE labs and helping students and faculty from underrepresented communities better access the science, technology, engineering, and mathematics (STEM) opportunities available to them at the national labs and beyond.
“Just advertising an opportunity is not sufficient,” said Brown, who has been involved in other diversity, equity, and inclusion programs, including at the University of California, Merced. “You have to make personal contact. The students need to be able to see themselves in this environment and see how it could work. There are a lot of barriers to getting people from untraditional backgrounds into the labs and universities, and you have to understand these barriers to build a program that will bring them in.”
A Unique Approach
These challenges have helped guide the form and function of the SRP program. Each year the program starts with faculty from these institutions applying to attend a "matching workshop," usually held in December, in which they meet with Berkeley Lab Computing Sciences staff, learn about what the Lab does, and explore what research opportunities might be available in the coming year. During the workshop, attendees are also able to speak directly with staff members, and slowly but surely, connections are made and potential research collaborations formed.
"Coming from a small liberal arts college (Hood College), SRP opened research opportunities to me on par with bigger universities and gave me a strong foundation for understanding the ins and outs of conducting scientific research. I also learned the importance of being able to communicate my research to an audience. Since the SRP program, I have been working with Dr. Crivelli to continue our research related to deep learning in proteomics and the Million Veteran Program, a collaboration between the DOE and the Veterans Administration to build predictive medicine tools for clinicians." -Rafael Zamora-Resendiz, Lawrence Berkeley National Laboratory, SRP 2018
“The matching workshop is the core piece, the beginning of the process,” Brown emphasized. The next step is finding funding to support the matches. Over the last decade, this has come from a range of sources: the Visiting Faculty Program, which provides opportunities for faculty members from institutions historically marginalized in STEM to enhance research capabilities, strengthen STEM education and learning practices, and develop talent to contribute to DOE research areas; the DOE's Science Undergraduate Laboratory Internships (SULI) and Community College Internship (CCI) programs; Berkeley Lab’s Laboratory Directed Research & Development (LDRD) program; the DOE Advanced Scientific Computing Research (ASCR) program; and a growing number of Lab staff members who participate in the program and contribute funds from their own projects to support the SRP teams.
For the past two years, the success of the program has led to new relationships with the DOE’s Exascale Computing Project (ECP) and the DOE Computational Research Leadership Council, which has enabled other DOE national labs – including Ames, Argonne, Brookhaven, Los Alamos, Livermore, Oak Ridge, Pacific Northwest, Sandia, and SLAC – to begin adopting the SRP program into their recruitment and workforce development efforts as well.
"My SRP experience laid the foundation for my career. I graduated from the University of Kentucky computer science master’s program with my Berkeley Lab work (developing and applying machine learning models on HPC infrastructures) as the foundation of my thesis. I then took a job at Lawrence Livermore National Laboratory as a chem/biodata scientist. I’ve been able to make many great connections that have helped me progress to pursue my Ph.D. at the University of California, San Diego, and I am very grateful for the opportunities that became possible through my experience with SRP/SHI." -Derek Jones, Lawrence Livermore National Laboratory/UC San Diego, SRP 2016/2017
“ECP approached us a couple of years ago and asked if we could create a program for them that mirrored SRP,” Leung said. “Since then, this collaboration has given us the ability to expand SRP and create a more cohesive and comprehensive program.”
“ECP is this massive multi-lab effort, which has made implementing a multiple-lab approach for SRP much simpler,” said Dan Martin, a member of the ECP leadership team and group lead for the Applied Numerical Algorithms Group at Berkeley Lab. “The structure was already in place, so we were able to leverage it to communicate across the labs and get more people involved and engaged.”
"I did not know much about the SRP program before I joined, and I was a bit anxious about an internship at Berkeley Lab. It hit some feelings surrounding imposter syndrome. But when I learned more about SRP through their website, I felt more at ease knowing I did belong and that I would be supported. I am now on a path toward graduate school and feel I have relevant experience that has prepared me to engage further in academic research." -Kristen Dawson, San Francisco State University, SRP 2022
In 2023, for example, there are 189 SRP participants spread over the 10 participating labs and 120 of the participants are being funded in some way through ECP projects, Martin noted. Eighty-one of the participants will be at Berkeley Lab. “It’s the largest cohort we’ve ever had,” he said.
“SRP has been more successful than we could have expected by a number of metrics,” Brown said. “Students have gotten major scholarships out of this, and there’s been at least one DOE early career award for one of the faculty members (Tanzima Islam). We’ve also been able to ensure that some of the SRP students and faculty are able to attend a professional conference or workshop and present results from their summer research.”
Life-Changing Impact
The program has had an impact on Berkeley Lab staff as well, Brown added. “They have come to see this as a very important mechanism for bringing in talent for our summer student program and beyond.”
"My experience with SRP kickstarted my entire career. I did not have a strong sense of how to navigate entering the research field and was having a hard time finding organizations that were willing to support early research students. SRP was a very receptive and supportive program that helped me feel prepared to enter the research community. The work that I did at Berkeley Lab has allowed me to attend conferences around the world to present my work, and now I am continuing my research on randomized numerical linear algebra this summer at another institution. SRP fundamentally changed my future and gave me the foundation to enter the world of research science." - Chris Camano, San Francisco State University, SRP 2022
Bert de Jong, Berkeley Lab AMCR’s Computational Sciences department head and deputy director for the National Quantum Initiative’s Quantum Systems Accelerator (QSA), has been involved with SRP since 2015 and has mentored some two dozen students over that time. It continues what has long been a passion of his: working with early career scientists to expand the pipeline of scientific talent that comes into the national labs and share the lab’s expertise in the process.
“The summer projects tend to line themselves up with popular topics, so the last couple of years that’s been quantum,” de Jong said. “We brought in students who had zero background with quantum, and we taught them about it. The same is true with machine learning; we taught them how to use the power of machine learning to accelerate their science.”
One SRP collaboration that stands out in his mind is with the University of Texas, El Paso (UTEP), which has sent student teams to Berkeley Lab through the SRP program under the guidance of Jorge Muñoz, an assistant professor of physics at UTEP. Muñoz and his students spent the summers of 2020 and 2021 working virtually with de Jong, studying machine learning methods to reduce the computational cost of density functional theory-based molecular dynamics simulations to enable the prediction of thermodynamic quantities of materials from quantum mechanics. Another success story is Samah Saeed, a computer scientist and engineering professor at City College of New York who spent the summer of 2021 at Berkeley Lab through the SRP program and recently received a DOE RENEW award to continue the quantum computing research she started at the Lab – now with de Jong as her co-PI.
“The impact [of SRP] is not just on the research but on the students and faculty having an opportunity to grow in their careers,” de Jong said. “And that is exciting to me because it means we did something right. Success breeds success.”
“This program has changed people’s paths by getting them to the Lab where they can envision themselves on this track and by exposing them to a possible career at the Lab or in industry,” added Martin, who also mentors students at Berkeley Lab each summer. “It really is about building new pathways.”
Dan Rosa de Jesús, who was part of a team from the University of Puerto Rico Mayagüez that came to Berkeley Lab in the summer of 2018 through SRP and was mentored by Crivelli, is one of many shining examples. The UPR team’s research focus was applying deep neural networks (DNNs) to protein structure classification and prediction using a new type of DNN, and their work led to a published paper and a presentation at SC18 later that year – exciting results for any young scientist. But the experience had a deeper impact for Rosa de Jesús as well.
"My time at Berkeley Lab through SRP was a unique opportunity to witness first-hand what the ‘big leagues’ of research look like and what was expected from a scientist at that level. Being able to present the team’s work at SC18 was a defining moment in my career as, through the questions and feedback from other researchers, I realized I had done something important toward resolving a challenging problem that will have positive implications for the development of medicines in the future.” - Dan Rosa de Jesús, Pacific Northwest National Laboratory, SRP 2018
While at SC18, Rosa de Jesús also learned about another opportunity: the GEM Fellowship. A few months later, he applied and was accepted as a Ph.D. intern at Pacific Northwest National Laboratory, where – after earning his Ph.D. from the University of Texas, El Paso in 2020 – he now works as a data scientist developing data augmentation methods for non-proliferation remote sensing applications. He is also “paying it forward” by mentoring younger colleagues and students and actively promoting a community of cultural inclusion as co-chair of the Hispanic Outreach for Leadership and Advancement Employee Resource Group at PNNL.
“Being at the right place, at the right time, and with the right people – all thanks to the SRP program – helped me direct my academic and professional career in the right direction,” Rosa de Jesús said. “I get goosebumps when I tell people about all the cool things I get to do at PNNL.”
About Computing Sciences at Berkeley Lab
High performance computing plays a critical role in scientific discovery. Researchers increasingly rely on advances in computer science, mathematics, computational science, data science, and large-scale computing and networking to increase our understanding of ourselves, our planet, and our universe. Berkeley Lab’s Computing Sciences Area researches, develops, and deploys new foundations, tools, and technologies to meet these needs and to advance research across a broad range of scientific disciplines.