Hosting a Hackathon during a Pandemic
When we were most isolated, we found creative strength in collaboration towards a common goal.
On March 26, just two weeks after the World Health Organization declared the coronavirus a global pandemic, the United States surpassed China, Italy, and every other country to lead the world in its number of confirmed cases (New York Times, 2020) — and the title has remained ever since. On the same day, over 20 teams of innovators across the world assembled virtually at the Global COVID-19 Relief Coalition Hackathon (goGCRC) to begin architecting medical, social, and technological solutions to the effects of this virus, which just over a month ago had been barely a blip on most Americans’ consciousness but was now rapidly re-rendering every feature of public and private life.
Hosted by the Global COVID-19 Relief Coalition, a fiscally sponsored project of The Giving Back Fund, along with scientists from Stanford University in partnership with Entrepreneurship at Cornell University, Weill Cornell, NEXUS, Summit, Amazon Web Services, Vocareum, and The National Institutes of Health, the hackathon brought together doctors, data scientists, lawyers, activists, and other innovators for sixteen weeks. Unlike a traditional hackathon, which might bring to mind images of caffeine-fueled programmers burning the midnight LEDs in a repurposed high school gym, the goGCRC Hackathon was designed to address problem spaces across the entire COVID-19 experience: not only working towards a vaccine but also filling cavities in our infrastructures around mutual aid, personal protective equipment (PPE), healthcare availability, and grief — among many others. Another difference, of course, was that it took place entirely online.
Utilizing Slack, Zoom, $180k of credits provided by Amazon Web Services, and a hosting platform, the Collective by RoundGlass, the GCRC hackathon successfully hosted over 20 cross-disciplinary projects from computer vision-automated mask fitting to virus data visualizations to an effort to better support families virtually grieving the loss of a loved one. Here are a few of the projects that were undertaken.
Predictive algorithm wearables study. Learn more about the initiative on their website.
Stanford Healthcare Innovation Lab and Snyder Lab’s COVID-19 Wearables Study
As the original project that sparked the idea for this hackathon, the wearables study conducted by Stanford Healthcare Innovation Lab aims to establish whether wearable devices can predict the onset of COVID-19 before symptoms appear. Wearable devices such as Apple Watches and Fitbits routinely collect metrics such as step count, skin temperature, and blood pressure, which have been used for early detection of other diseases. Researchers explored the applicability of this data to COVID-19. The Stanford researchers were able to detect physiological markers associated with COVID-19 in 80% of infections, and up to 8 days before symptom onset. You can view the preprint of the study’s results on Medrxiv. Due to the confidential nature of the clinical trial, additional hackers did not join this initiative. However, the GCRC infrastructure supported the team with connections to valuable partnerships and resources including funding.
Grieving in a Pandemic: Funeral White Paper
The world is experiencing death at a magnitude and intensity that are, for most people, unprecedented. 973,000 people worldwide have died from COVID-19, each leaving behind loved ones. Rites like funerals and family gatherings can take place only virtually, if at all. Photos circulate online of bodies piled into trucks. Unique to this pandemic, many family members face feelings of guilt for having spread the virus to their loved ones who have passed.
The Virtual Funeral Collective, a group of over 70 doctors, nurses, grief counselors, psychologists, academic grief researchers, and funeral home directors from around the world came together at the GCRC hackathon to develop resources to help people cope with death and grieving. The Collective’s work culminated in a white paper covering topics ranging from complex emotions, safely finding connections, and maintaining the sanctity of a loved one’s passing during this time. Their work stems from an understanding of grief not as merely an unpleasant emotion to be discarded as soon as possible, but as lifelong labor that “lasts as long as love lasts.” The Paper’s resources are extensive. While the paper is targeted at healthcare workers and others that guide grieving individuals or deal with death and dying at work, it includes many practical suggestions that can be utilized by individuals as well, including ideas for holding socially-distanced memorials, virtually supporting grieving individuals, and talking about death in a way that is direct and meaningful.
Open COVID Pledge
One of the initial goals created for the GCRC hackathon was to identify novel ways to increase testing, and one idea that arose from that goal was “Free the IP,” a campaign to make intellectual property freely available for use against the pandemic. GCRC members teamed up with lawyers and scientists at Creative Commons, Stanford University, UC Berkeley, University of Utah, DLA Piper, Helpful Engineering, and the University of Cambridge to create what ultimately became the Open COVID Pledge, an initiative that is now led and stewarded by Creative Commons.
Open COVID Pledge recognizes the moral and practical imperative for every available resource to be deployed for ending the COVID-19 pandemic and mitigating its effects. While the sharing of intellectual property is often limited by licensing laws, lawyers and scientists across seven universities and organizations created a patent pledge that would make coronavirus-related research and innovations freely accessible to use without impediment.
Highlights of pledged IP and a database hosted by IPscreener to search through all the pledged intellectual property are available on the Open COVID Pledge website. The pledged projects are myriad as those undertaken by the GCRC itself. AT&T pledged a patent that optimizes an ambulance route to transport a patient to a hospital as quickly and safely as possible. NASA JPL included designs, data, and instructions for 3D-printing respirators. Sandia contributed multiple patents, from a safer drive-in testing booth to a method for decreasing false positives in COVID-19 testing. Other prominent pledgers include IBM, Mitsubishi Electric, and Intel.
Zanzibar is an archipelago off the coast of Tanzania, home to 1.3 million people. Zanzibar functions as an independent entity with its government and independent healthcare system and has three hospitals, two of which are directly run and managed by Health Improvement Project Zanzibar (HIPZ) a UK-based nonprofit. In 2006, at a time when there were zero doctors or nurses across these hospitals, the Zanzibari Minister of Health invited Dr. Ru MacDonagh, Chairman of HIPZ to help restructure their healthcare system. Since then, HIPZ has created a thriving healthcare system that offers prenatal and postnatal interventions to reduce infant mortality, mental health, pediatrics, dental care, and even surgery. They first brought over doctors from the UK and then developed a system to train and engage local healthcare workers, and now support over 300 healthcare workers.
Before the pandemic started, the Stanford Healthcare Innovation Lab had been talking to HIPZ about the possibility of collaborating on healthcare innovation research that HIPZ is conducting in the areas of nutrition, diabetes, and patient care. At the start of the pandemic, when the Global COVID-19 Relief Coalition asked HIPZ whether they had any problem for hackathon teams to solve, they told us that they did not have enough or proper PPE to protect their healthcare workers. Because Zanzibar has so few healthcare workers, and a limited ability to train more, the loss of any health professionals could mean a continued loss of care for hundreds of thousands of people.
On paper, there exist many challenges in creating a strong virus response on this island. Social distancing is virtually impossible as most households are large and multigenerational. Their hospitals lack the sterilization equipment needed to reuse N95 masks, and they didn’t have access to any. Large religious gatherings are of vital cultural importance, and health issues such as diabetes and hypertension are common. Many people told this team that even attempting to aid Zanzibar was futile.
Several groups of people started thinking up different solutions with guidance from HIPZ, including trying to get local tailors to follow designs, using solar UV to sanitize masks, and creating plans to inform COVID-19 public health response. As we brainstormed solutions, HIPZ contacted us to provide an update, letting us know they were running out of masks, forcing doctors to see suspected COVID patients without masks, and forcing them to purchase more at high local prices on their already-thin budget. Their ability to support ongoing healthcare was drastically reduced.
Ultimately, the group approached several mask suppliers, and the Tach Group, a growth and strategy partner for global brands that began supplying masks during the COVID-19 pandemic, donated 100,000 L1 masks. Private Medical, a concierge medical service in the Bay Area stepped up and supported the cost of shipping the masks to Zanzibar. Global Health Crisis Coordination Center (GHC3) and Flexport and helped with logistics. After delays around canceled flights, the masks arrived in Zanzibar on September 11th.
Overall, this project was proof that an effective COVID-19 response is tailored at a grassroots level. Working with local changemakers allowed us to help to meet them in a way that could make a difference and meet actual on the ground needs and support the work that HIPZ is already doing.
These masks will allow healthcare workers to continue their vital work. The team is also looking ahead at how they can continue to provide Zanzibar with both disposable and reusable masks.
Texas Children’s Hospital x Skylark Labs: AI Mask Fitting
Skylark Labs, an organization that creates artificial intelligence solutions, took on a project to use computer vision to automatically fit healthcare workers at Texas Children’s Hospital with masks so a manual fit-test does not need to be done for each individual.
The high cost, imprecision, and resource shortage associated with manual fit tests spurred Skylark Labs’ vision for a fully-automated process that first detects whether an individual is wearing a mask, then checks if it is being worn correctly. This would allow fit updates to occur more frequently than once a year without draining the hospital’s resources — which is important, as facial changes such as injuries and dental procedures can routinely affect the sizing of the required mask.
The team gathered data from five hospitals in India; this data is being fed into a deep learning pipeline that should begin to “learn” the process of shaping and sizing face masks to, thus, reserve hospitals’ valuable time and resources for COVID-19 prevention and treatment.
Ultimately, Texas Children’s hospital did not end up using the algorithm, but it is being piloted in several contexts in India to detect whether people are wearing masks and support social distancing. As with pandemic solution-building and as with mask fitting, one size does not fit all.
These are just five of the 27 extraordinarily innovative projects that aimed not only to end this pandemic but also to sustain us as we learn to live through it. The cross-disciplinary nature of these projects was by far the most successful aspect of the hackathon; the strong targeted outreach to individuals within this larger framework was effective as well.
What was harder, though, was expecting this large, virtual community to engage and communicate with each other. Communication and team updates proved particularly difficult to get. The Collective, our hosting site, was not fully utilized by project teams. To adapt, the organizers had to be comfortable not receiving project updates from the teams and trust that if any team needed assistance they would reach out. What ended up working, however, was taking the opposite approach: not imposing a workflow on each team, but allowing them to organically create their own. While many projects and teams dissolved or did not complete projects, the meaningful projects that did emerge had a real impact.
These hiccups, however, only represent a learning curve for future hackathons and remote environments as we continue living in this virtual world for the next several months. Whether in a research lab or our own homes, we are all being called to adapt as deftly to this new world order as this virus has adapted to us. This hackathon represents not the end, but the beginning of that innovation and resilience.
Thank you to
for help writing and editing as well as overall mentorship.
The Hackathon was organized by Ami Stuart, Rose Pember, Tushar, Nikki Solanki, Tanya Chetri, Joi Ito, Ariel Ganz, Benjamin Rolnik, and myself ( Sarah Wilen), with help from Michael Hebb and Maya Hough at the Collective, as well as guidance from Rudy Tanzi, Jim Doty, Kayleigh Millet, Natalie Frank, Mark Radcliffe, Allissa Dillman, Marc Pollick, and Michael Snyder. Thank you to Amir Bahmani and Arash Alavi for building the PHD Dashboard.