Managing COVID vaccine distribution in West Virginia | RStudio
With a little help from open source software Learn more: Data Driven West Virginia: https://business.wvu.edu/research-outreach/data-driven-wv DDWV PPE forecasting: https://wvutoday.wvu.edu/stories/2020/04/27/wvu-business-experts-partner-with-the-national-guard-to-forecast-ppe-needs DDWV inventory management system: https://wvutoday.wvu.edu/stories/2021/03/22/a-different-kind-of-science-wvu-chambers-college-data-scientists-propel-west-virginia-s-acclaimed-vaccine-strategy-with-digital-inventory-management-system West Virginia National Guard: https://www.wv.ng.mil/ Shiny: https://shiny.rstudio.com/ West Virginia leading nation at start of vaccine rollout: https://www.vox.com/first-person/2021/3/4/22313540/covid-19-vaccine-west-virginia *** In the United States, approximately 2.5 million doses of COVID vaccines are being delivered each day, and how these doses go from the manufacturer to a shot in someone’s arm varies by state, often with mixed results. But early on in the vaccine distribution process, one state led the pack in terms of using the majority of vaccine doses it had been allotted. That state? West Virginia. Part of what has made West Virginia successful is the creation of an inventory management system using Shiny, an open source framework for building interactive web applications. The system was built by Data Driven West Virginia, part of the John Chambers College of Business and Economics at West Virginia University, in collaboration with the West Virginia Army National Guard. Using Shiny has provided visibility into each component of the vaccine supply chain, leading to the creation of distribution plans that are able to quickly and efficiently match supply with demand, getting vaccines to the right people in the right location at the right time. *** RStudio’s mission is to create free and open-source software for data science, scientific research, and technical communication to enhance the production and consumption of knowledge by everyone, regardless of economic means, and to facilitate collaboration and reproducible research, both of which are critical to the integrity and efficacy of work across industries. RStudio also produces RStudio Team, a modular platform of commercial software products that give organizations the confidence to adopt R, Python and other open-source data science software at scale, along with online services to make it easier to learn and use them over the web. Together, RStudio’s open-source software and commercial software form a virtuous cycle: the adoption of open-source data science software at scale in organizations creates demand for RStudio’s commercial software; and the revenue from commercial software, in turn, enables deeper investment in the open-source software that benefits everyone. Check out www.rstudio.com Follow us on Twitter: https://twitter.com/rstudio Facebook: https://www.facebook.com/rstudiopbc/ And LinkedIn: https://www.linkedin.com/company/rstudio-pbc/
image: thumbnail.jpg
Transcript#
This transcript was generated automatically and may contain errors.
When you're in the weeds and you're just coding and looking at the UI, you don't think about this is a life or death situation.
When the vaccine stuff came around, I knew, do we have the right inventory at the right place Distribution in West Virginia is managed by the National Guard, predominantly. You know, it wasn't as efficient as we would like, and we were aware of that, but you know, we're in a pandemic, there's no playbook here. There were a lot of hands in different files, you know, they were working with Excel files. I knew we needed an advanced planning system. And when major costs hit the ground, he shot me a text and he said, hey, we need a tool. It wasn't even a question, of course we're going to do that.
Building the distribution tool
Distribution in West Virginia is managed by the National Guard, predominantly. The project started with just giving them a more automated process for different groups to submit the data that ultimately gave them the information that they needed to develop their distribution plan. In our state, we have five distribution hubs. If you look at a map of West Virginia, you can basically touch the four corners of the state, and you're going to find a hub, and then there's a hub in Charleston.
The key priorities were simplicity, right? I mean, we got to have something where at hub level, at user level, they just have to log in, key in a number and hit enter. And we kind of looked at each other and said, we don't have time to build a database. We needed something that was cost effective, but that we could get it up and running quickly. We knew we'd get to a place where they wanted more visualization. These are government entities. There's not an open-ended checkbook. This is taxpayer money. We're like, what can do all of that for us? So Shiny just became the quick and easy solution that came to mind for both of us.
So Shiny just became the quick and easy solution that came to mind for both of us.
We needed to create something that would ingest, had about 10 different people submitting every week, and then aggregating that information into an output that would allow them to make strategic decisions. The manual way, at best, we would get to that finished product Sunday night and have to start operations Monday morning. So we're now, we have that ready to go by 4 p.m. on Wednesday.
Real-time visibility across hubs
The reason we've been able to be successful, I think, is because of this distribution network, knowing where these vaccines are going at any given point. You're going to run into some problems, right, where you may not have the right quantities necessarily at the right place. So that does require some, what we call, inner hub transfer, so to speak. Based on the tool we built, we're able to provide visibility clear through this project. And so we're seeing real-time draining of supply from the hubs.
In combat, I'm moving fuel, ammunition, medical supplies, maintenance repair parts. And there's this thing called the enemy out there trying to take you out while you're doing it, right? I've always taken the approach of if you can do it in that environment, you can do it in any environment minus the enemy. Sometimes I have to tell myself, like, no, it's actually a huge feat that we've accomplished all of this, because it just seems like we've all been doing our jobs. The entire principle behind this is build something in an agile manner that you are going to repair day. And the happiest day of this entire thing is going to be when you turn it off.
And the happiest day of this entire thing is going to be when you turn it off.
