Using Data to Build a Better Future: An Interview with Dr. Andrew Schroeder
Tell us a little about yourself and your research.
So in my professional life, I actually wear a couple of different hats. Primarily, I’m the Vice President for Research and Analysis at Direct Relief, which is one of the largest global humanitarian aid organizations in the health sector in terms of the value of aid delivered. We also work throughout the United States supporting community health centers with access to essential medicines and medical supplies. I’m also co-directing a collaborative project through Direct Relief and Harvard University’s schools of Public Health and Medicine called CrisisReady, which focuses on improving how researchers and response agencies use human mobility data linked to epidemiological analysis for global health emergency response. I’m also the co-founder and board president for a non-profit organization called WeRobotics, which helps to create and sustain local capacity in the use of robotics to address key problems in disaster relief, health, environmental protection, and economic development. My work across all three of these areas focuses on how we can use spatial data and analytic methods to respond to emergencies and improve the health and well-being of communities.
Your work, in part, focuses on research and analysis for humanitarian relief. What are the most recent trends you’ve seen emerge in this field?
Just to pick a couple: 1) There is I think increasing interest throughout the humanitarian community in localized responses — meaning that people in disaster-affected areas are themselves leading the response actions — which means that those folks need access to the kinds of data and analytic tools that have for some time been concentrated in headquarters. “Who” is doing the work matters as much or more as what work they are doing, and we have to be able to support that positive development, 2) There has been this really rapid growth over the past 10 years in the number of people globally using mobile devices, which produce data on where people are, how they move, and what sorts of risks they may be exposed to given their locations. We saw this in a huge way during the pandemic when distancing policies and other measures were imposed and this really high demand was expressed throughout the world for data to monitor those measures. This is really tricky data to use effectively though, and especially in a way that fully respects privacy and community rights to protection. It’s also just large scale and rapidly changing, which poses lots of challenges for resource-poor organizations that already struggle with more basic data analysis. We’re seeing a lot of groups trying to figure this out, and there will be several important developments in that field coming up shortly.
What are some of the most useful ways in which data scientists can address these trends?
In terms of trend #1 — localization — I really think that data scientists can voluntarily contribute time to organizations which need work done in many different parts of the world. There are established models for this — groups like GIS Corps which puts people with geospatial skills to work on relevant global projects for instance — but regardless, it makes sense to find ways to assist with skills transfer and training wherever possible. In terms of trend #2 — big mobile datasets — I think there’s a real need out there for much more effective data management to help corral these datasets into units that can be more easily analyzed by more people. A friend of mine posed this as a kind of “Data Engineers Without Borders” which would focus specifically on the un-sexy but absolutely essential work of organizing the world’s data into analysis-ready form. That could be a big opportunity to make an impact I think.
What specific challenges do researchers face when applying their research to direct relief?
I think one of the key problems is just conveying the context of what we’re working on effectively to folks who may be able to help, and putting our problems into step-wise structures that might let people engage on different but related portions of a larger issue. Here’s one thing we really need: forecasts of unmet medical need for vulnerable communities. How can we create “market” forecasts for places where the market itself is failing, in order to understand how to meet effective demand for medical goods that are intended to be given away, not sold? It’s a tough problem — and it requires a reasonable amount of contextual awareness and understanding to solve.
What more needs to be done to improve the application of data science to direct relief?
So much … but to name just one: it would be amazing to have a kind of library of Python and R code that’s tailored specifically for analyses faced by humanitarian organizations … which build on open data sources and clearly defined problems … so that the broader data science community can help us and other organizations to expand a scripting archive which would be really challenging to building exclusively with our own resources. If there are folks out there that would like to contribute code to this kind of open humanitarian archive they should get in touch!
Your work at WeRobotics.org also focuses on building robotics applications for humanitarian aid. What are some ways in which broader access to robotics can be achieved?
I think the fundamental insight of WeRobotics is that the world is absolutely full of astonishingly talented people who are eager to get involved in building a better future where we can use advanced autonomous systems to solve community level problems. They just need to be identified, and connected, and resourced, and sustained with supportive communities of practice through like-minded peers. So that’s what we do — build those connections, make them stronger, direct training, contract and job opportunities, and other resources towards them, and then mostly get out of the way as they take the lead. Of course … doing that also means helping to educate much of the world about the difference that “who” is doing the work makes to successful projects.
What resources would you recommend for anyone interested in learning more?
If you’re interested in learning more about Direct Relief you can find us at https://www.directrelief.org. If you’re interested in learning more about large-scale mobility data for aid and health check out our resources at CrisisReady https://crisisready.io. And if you want to learn more about WeRobotics take a look at our website at https://werobotics.org
To learn more from Dr. Schroeder, sign up for his free webinar, Women Displaced-Using Gender Disaggregated Mobility Data to Study Disasters, here.
