A global community of citizen scientists saved the MilkyWay@Home project. The project’s scientists lost their research grants two years ago amid a crisis in American astrophysics funding. The MilkyWay@Home community’s financial support - and personal sacrifice from the project team - kept the project afloat. Then last Fall the National Science Foundation awarded them a three-year research grant.
Dr. Heidi Newberg is an astrophysicist at Rensselaer Polytechnic Institute and the principal investigator for MilkyWay@Home. I interviewed her in 2014 and again in 2015 for articles about the project's fundraising campaigns. Now, with the project’s future secured, Newberg described the importance of the volunteers’ support.
“The community support over the past two years made a huge difference,” Newberg explained. “As you know, I lost funding for MilkyWay@Home two years ago, and lost all of my federal funding one year ago. If the community had not supported me, I would not have been able to keep everything running, and that means I would not have gotten the grant I have this year.”
Searching for the Milky Way’s dark matter
MilkyWay@Home has ambitious goals to model the dynamics of our galaxy’s halo. Beyond the galactic core, beyond the spiral arms, the Milky Way’s halo is a sphere of stars and gas and dark matter. Many of these stars were once part of other galaxies shredded by the Milky Way into tidal streams. Newberg’s team creates computer simulations of this billion-year process.
Once, a project like this would have required its own dedicated supercomputer. But MilkyWay@Home uses the power of crowdsourcing to get the job done at a fraction of the price. It uses distributed computing to divide the work into small chunks.
Twenty thousand active volunteers let their personal computers crunch the numbers. MilkyWay@Home’s servers collect and combine the chunked results. The volunteer community, in effect, forms a global virtual supercomputer.
“There are two MilkyWay@Home projects that are very close to producing important astrophysical results,” Newberg explained. “One project measures the spatial density of the Milky Way halo (including large tidal streams of stars ripped from dwarf galaxies as they fall into the Milky Way).”
“The other project is a proof of concept that we can measure the dark matter content of what was once a dwarf galaxy, but was ripped apart by the Milky Way's gravity so that all that is left is a tidal stream of stars.”
A few years ago, however, MilkyWay@Home almost came to an end. A perfect storm had hit America’s astrophysics funding system. The construction costs of cutting-edge observatories rose just as a series of shutdowns and automatic spending cuts hit government budgets. The NSF had to slash its grants programs.
MilkyWay@Home was one of many casualties. The NSF declined to renew its grant for the 2014/2015 academic year.
“Without funding,” Newberg told me at the time, “it would be hard or impossible to keep up operations.”
Sacrificing for science
Newberg took desperate measures to keep the project alive. She gave up her share of the 2015 Breakthrough Prize in Fundamental Physics. She even worked at half pay during her sabbatical. Times were even more desperate for Newberg's students. They depended on the research grants for their annual stipends.
Jake Weiss is one of the MilkyWay@Home graduate students. “The financial struggles have made working on this project extremely difficult and stressful," he told me in 2015. "I am plagued by the fact that we could run out of funding and I may not be able to complete my PhD.”
Looking back now, Weiss said, “life as a grad student was rather difficult. I did not have much of a life. As a first year physics PhD student, you are required to take very difficult and demanding classes. I was also responsible for helping to secure funding for the group and myself [which] took significant time away from doing my class work and research.”
Citizen science to the rescue
For most researchers the loss of funding leaves only one choice: shut down the project. But Newberg had something that most researchers do not. A global community of volunteers backed her research.
More than 200,000 people have contributed their computers’ spare processing power to MilkyWay@Home.
Newberg launched a fundraising campaign during the 2014/2015 academic year. It raised enough to keep the project going with a skeleton crew. Another fundraising campaign ran during the 2015/2016 academic year. It raised enough to keep MilkyWay@Home going through the summer of 2016.
“Particularly during the year with no federal funding,” Newberg recalled, “I relied primarily on gifts from private donors to fund my students. The private funding also helped me to attract some help from Rensselaer Polytechnic Institute.”
But even with that help, the future looked uncertain.
“By the time the Summer came around,” Jake said, “Professor Newberg and I discussed what we would do if we did not receive NSF funding. It was day to day at the end of August trying to determine where my funding was going to come from for the Fall.”
All systems back to normal
The reprieve came in early September. The MilkyWay@Home project received a three-year, $350,000 NSF grant.
“[It] ensures I can finish my PhD without the uncertainty of yearly funding campaigns and the large time commitment of a teaching assistantship,” Weiss said. “The more time I can spend on my research, the more likely I am to get meaningful results in the limited time I have left to work on this project as a graduate student.”
The grant also relieves the burden on MilkyWay@Home’s volunteers. They already do much more than contributing extra compute cycles.
“The MilkyWay@Home volunteers have given us a very large and valuable computing resource,” Newberg explained. “In addition to that, we have received other really important gifts from the community including teaching us about GPUs when they first became available and helping us trouble-shoot the software system.”
“They expect us to keep our side of the bargain, though. We are responsible for sending them as many work units as they can handle, that take a reasonable amount of time to compute, and assigning a fair number of credits to each of those work units.”
Weiss agrees that attention to the community is critical. “Having a community in a citizen science project is about ensuring good communication between project staff and volunteers,” he advises. “It is also about ensuring your volunteers feel appreciated for the important work they are doing and trying to get them to be invested in the future of the project.”
Pete Boulay is one of the project’s participants. He is also a volunteer moderator, under the handle Blurf, on the MilkyWay@Home forum. I asked him recently to explain the community's dedication to Newberg's project. “The team, as a whole, has increased their visibility and responsiveness to the crunchers,” he replied. “We all do our best to answer questions and address issues as quickly as possible. This base level interaction is CRITICAL to a project that relies on this type of work.”
Newberg said that Boulay “lets me know when the community is unhappy, gives me advice when I am unsure how the community will respond, makes sure people are behaving on our forums, and is a huge help managing the citizens in the citizen science context. Thanks go out to Blurf for helping to keep the volunteer community together, and for significant help in fundraising.”
Now that the funding crisis has passed, Newberg can boost the MilkyWay@Home staff and its research output. Now her students will get to conduct the cutting-edge research that will launch their scientific careers.
“I would like to thank our current community for their help and support,” Weiss said, “especially over the last couple years where we were close to shuttering our project due to lack of funding. It was truly amazing to see how generous our community is.”