 Welcome to MIT Supply Chain Frontiers, brought to you by the MIT Center for Transportation and Logistics. Recently, food security and food systems have become top of mind as we face seasonal disruptions and unusual disturbances caused by the pandemic. MIT CTL's Tim Russell speaks with his fellow humanitarian supply chain lab researchers about systems mapping and how they're using rapid ecosystem assessments to inform government and the private sector in diverse scenarios. Take it away, Tim. Welcome, everyone. Today we're speaking with the humanitarian supply chain lab contributors about their innovative systems work, responding to emergencies, specifically around COVID this last few months here. Today I want to welcome Dr. Jared Gensel, the founder and director of the Panitarian Supply Chain Lab. Thanks, Tim. It's going to be here. Chelsea Graham, a graduate of the MIT Supply Chain Program and a research engineer in the Humanitarian Supply Chain Lab. Thanks, Tim. Great to be here. And Tristan Downey, a graduate student in the technology and policy program here at MIT. Thanks, Tim. Before we go into depth on the methodology that we used to respond to this emergency, I wanted to ask for a brief description on the projects that Tristan and Chelsea were working on. Tristan, can we start with you? And can you tell us a little bit about the rapid systems assessment that you recently led in Uganda? Yeah, we did a rapid system assessment for the agriculture market system in Uganda. Doing this, we basically surveyed about 300 separate sources of information for what was going on on the ground. From that, extracted hundreds of distinct facts and layered them onto a system map, from which we determined key impacts on the system and what that really means for specific actors on the ground. Great, thanks, Tristan. Chelsea, could you describe a little bit of the work you did with SCAN? Yeah, thanks, Tim. So we were contracted by FEMA to provide ecosystem assessments as part of the supply chain analysis network. And what that entailed was similar to what Tristan did. We tried to provide rapid assessments based on what we saw happening in the market in terms of food supply chain within the US. And we also relied on a Sentinel network of industry partners to ground truth that information. And our goal was to let FEMA know if there were any bottlenecks occurring within the food supply chain. And if so, what type of action would they need to take to remedy that? Thanks, Chelsea. Next, I want to turn to Dr. Gensel and ask a little bit about the methodologies that both of these two projects use. Thanks, Tim. It's two methodologies that came out of different places when we started seeing linkages between the two. So we had already been before COVID starting to think about private sector supply chains from a systems perspective. So we started doing this work with FEMA and it's based on engagements we've had with them over the years and actually was formalized a little bit through work with the National Academies of Science, Engineering, and Medicine in understanding the systems behind private sector supply chains and how they adapt and respond in a crisis. Rather than just try to anticipate dynamic events as they happen, let's see if we can understand how the system works so that we can anticipate the adaptations and the shocks that happened during a crisis. And so then that just naturally played out during the engagement with FEMA. Completely in parallel, we were working with a different part of the US government with USAID in Uganda over the last several years looking at market interventions. So USAID trying to support the private sector in providing goods and services to smallholder farmers. It was really around agriculture in Uganda. So we had already been thinking more deliberately around systems and markets as systems in Uganda. So as we were talking with USAID about what we were doing in the US looking at grocery supply chains and making sure that grocery shelves can be filled, they were also saying, well, we wonder that the crisis wasn't as bad in Uganda but they were wanting to anticipate, well, how would markets change, private sector markets change in Uganda as well. And we hadn't really thought about trying to put those methodologies together but we've started exploring the work we did in Uganda with agricultural markets and contrasting that with the emergency supply chains of private sector supply chains for grocery stores in the US and started seeing some connections with the system's lens. Could you say a little bit more, Jared, about reaching out to trusted partners and people to ground truth this information? I know Chelsea mentioned ground truth, I mean, and Tristan mentioned talking to like over 300 different sources. Can you maybe tell us a little bit more about how that fits into this? The work in Uganda is inherently working with activities who are working with actors directly trying to support their work. And so it's obviously they're trying to facilitate so there's a direct connection on an ongoing basis. The work with FEMA, you need to get quicker reaction time. You don't have time to learn like we have been in Uganda over a longer period of time. You need to be able to in a crisis, you need to adapt quickly day to day. And so with the FEMA work, we try to think about something that's a bit more nimble than the effort it might take to learn about a system over time. And in 2017, we started with this concept of if we understand the structure of the supply chains of private sector supply chains, maybe we could identify key nodes in these demand and supply networks that could be ones to pay more attention to. So in the US, for example, this could be a grocery distribution warehouse which may not normally be thought of as a critical infrastructure component. I think of critical infrastructure being power plants and refineries and things that you would assume that are normally covered by what people tend to think of in the emergency space as critical infrastructure. We started looking at the whole supply chain and saying there may be other points where you need to pay attention. We actually learned in our work with FEMA that the refineries are monitored very closely but the fuel terminals where they actually load the trucks with fuel to go to the retailers is not monitored as closely. So this whole concept, looking at these complex networks, how can we anticipate where we would wanna have conversations on a more rapid basis? Because you have to work quickly. Who would we wanna talk to to assess what's going on and then be able to adapt quickly? And so that's where the concept of sentinels came in. Points where you try to monitor the health of the community, of society by taking regular samples and checking in on a fairly rapid basis. Normally that's in the sense of public health where you're trying to anticipate the flu season. So you'll pick a few key hospitals and be gathering really rich data on the nature of patients coming in and what symptoms they have and you try to track early if the flu season is starting to spike or not. So we thought, well, we can apply this maybe to supply chains and identify key points in these supply chains and therefore the key people who understand what's going on at those key points in the supply chain and interact with them. And that's where that concept of a sentinel network came in. We're literally talking to the people who are watching the supply chain as it's operating and saying what's happening. So that's what that sentinel network came about in FEMA. Thanks, Jared. Tristan, as Jared was just mentioning the sentinel network, maybe you could talk a little bit about how already having an understanding of the you got an agricultural market system was able to help you as you did a rapid assessment for COVID. Yeah, for sure. So we're taking a look at the whole system, right? So by mapping it out first, we can sort of start to place where all these bits of information are coming in from and by information sources, I mean everything from news articles to discussions we've had with USAID partners with people within USAID but also reports from other actors. So one thing that we found was that within the supply chain for, well in this case it was mainly maize, we found we had a lot of information about say importing and manufacturing. So upstream in the supply chain for agriculture inputs. And then we also had a lot of information sort of downstream. So we had a lot about consumers but also about say exports and traders and retailers. But one thing that we didn't have as much information about was about farmers themselves. So it's sort of the center of the supply chain there. So we sort of noted that that was what we'd call an information gap in the system. So we actually then went out to interview basically several farmer cooperatives to get more information there. But in addition to some key government statistics and surveys about other bits of the supply chain, we also set up interviews with key informants in those other parts of the supply chain as well. So importers and commodity distributors basically. So I guess having that initial system map in place really allowed us to make sense of information that was coming in and understand sort of where it had impact in the system as opposed to just seeing, oh, you know, maize price has gone up somewhere that means something. Well, no, it actually means something quite specific. And if we tease out how that price has gone up, we can see, oh, it might be that specifically just the retailer is making a higher margin now and maybe they're trying to recoup losses from before. But that price increase might not be going all the way through to the farmer. Thanks, Tristan. Chelsea, maybe I could have you comment on maybe some examples where this systems knowledge that Tristan's talking about, you're able to use that to get a better understanding of what was going on in the US. We took a slightly different approach in that we really honed in on what was happening with the freight market in the US in order to try to get a pulse on where bottlenecks might be occurring, is freight flowing or being stalled in certain areas. And this is really due to just not being able to analyze at the industry level across retail and food. So trying to get a holistic picture, we decided to focus on freight and freight included rail, included trucking, also shipping as well. We fortunately have industry partners that are in those various categories and they might be either a retailer that moves product with all of those modes of transportation or it might specifically be someone that's within the trucking industry that we could reach out to to get different perspectives of what's happening. So some perspectives are very product oriented while other perspectives are more related to freight movement itself. And that allowed us to create a really a well-rounded picture of what was happening in terms of how freight was moving across the country and how it was, we can't just look at food supply chain in a vacuum because the food that's being moved is those trucks are also moving other products across the country. So you really need to be able to see how they're all interacting with each other. Chelsea, do you have any examples of these interactions that were surprising? Yeah, one thing that I think everyone will remember in the media and some big headlines was related to the meat industry and meat packing. So there were a lot of COVID outbreaks within meat packing plants. And there was fear that there was going to be meat supply shortage due to this issue. So the COVID outbreaks caused the plants to shut down. And then that started to impact the upstream farmers that needed to get their animals to these plants. What we noticed was that while the plants were being affected because they were having to shut down, there was still a large amount of frozen meat being stored. And this did not actually impact the supply of meat that was going out into the country. Because we were looking at cold supply chain, we were talking to people in the industry that move the meat and had a good understanding of where this frozen meat is being stored. We were able to connect the dots to see that while the supply might have been impacted at the meat processing plant level, it was not actually going to impact the consumer. Tristan, let me give you a chance to say, did you have any unexpected findings as you were doing your analysis? Yeah, I guess so one interesting thing. Well, I guess some interesting themes were just differential impacts, so how it might impact different actors. And one was about exports actually. So the Ugandan government imposed some pretty strict COVID movement restrictions and border closures, but still allowed shipment across the border. But to do that, you need some certifications basically. So what we found was that although there were some increases in delays for formal exports, so those would be bigger companies that would have the capacity to get the certification needed to ship product across the border that completely halted informal trading. So that's something that is typically much smaller traders, often women in these border districts. And it's quite striking actually the export stats. Basically the formal export stats are not that far off of seasonal trends, but informal exports. As soon as that border closure came into place, they dropped to zero immediately and they've stayed at zero. And in terms of value for Maze alone, that's since COVID, that's millions of dollars of value lost. What did that mean for the farmers that normally do this informal trade? So that's a significant loss of income for these small informal traders who would be trading just across the border, just into South Sudan or into Kenya or into the DRC. So that was a big impact that wouldn't be immediately obvious just from looking at sort of normal aggregated export statistics. And as far as that informal export volume has stayed at zero for all commodities. So that'll have a disproportionate impact for people who are already vulnerable. These are already small traders who would already have less of an access to finance or other sort of cushions basically. Well, just to add to that, it disproportionately affects women because we also were able to pull together data from working in the system to see that these traders are often women, these informal cross-border traders are women. So yeah, just trying to piece together information from various sources. And I'll have to admit, Tristan developed a pretty interesting approach of taking facts from news feeds and different experts and synthesizing and pulling it together using our system map. And so there's a nice formal way to track everything in real time so we can connect those dots. No, that's interesting. And we saw the same thing in the US when we're talking about differential impacts to small businesses versus large businesses. Jared, you mentioned reaching out to different sources. Maybe you could do a compare and contrast with like the kind of people we're talking to in these two different projects, how often the frequency just give a little sense of how to reach out to these Sentinel networks is different between these two different approaches. We know that supply chain management is a human endeavor, it's human management. But when I was first thinking of these Sentinel networks, I was thinking of data feeds, like let's get Sentinel points, let's identify the key nodes and let's get data and in real time we'll get data and we'll do all this analysis and we'll see what's going on. And especially this year, we've kind of been learning over the years, but especially this year we realized that it's not just the Sentinel points and data about what's happening there, it's the Sentinels themselves. The people that are managing those critical nodes have information. And in a crisis, they actually want to talk, they're trying to make sense of what's going on at the same time. And especially something that's as unique as a pandemic. I think sometimes in the U.S., these grocery supply chain providers, then the transportation providers and so forth that we typically work with, hurricanes are somewhat normal and they're looking for new information, but in this pandemic, they're all looking for new information, they wanted to talk. So we had a much richer discussion. In fact, what ended up happening, I've called this, we had an industry peer review process for all of our assessments. We would do an assessment two or three times a week. We would send them out with that frequency and these industry people would read them and give us feedback. It's real time review with all of these different experts, these Sentinels across these important supply chains, getting information from them in real time to really get a sense of what's going on that probably would be, if we tried to wait for data to come in, it'd be too slow. So that rapid interaction and the desire of groups of people to want to learn together and make sense of what's going on together really helped connect. But one thing that's different in Uganda than US, I mean, US has very concentrated supply chains. When you're talking about food distribution, there's a few key food distributors that cover most of the grocery sector. And so if you talk to a few of them, you get a really good sense of what's going on. Chelsea mentioned before the meat packing industry, we could talk to a few people and get a pretty broad sense of what was happening. And Uganda, the supply chains are much more fragmented to try to understand what's going on in Tristan's case with cross border trade, there's hundreds of traders you'd have to talk to and there's not a really good consolidated place to do that. So that interaction is more challenging in that kind of situation. At the same time, that fragmented decentralized structure in some ways, they're all adapting themselves rapidly. They're talking to each other, they're figuring out how to adapt. And I think one of the gaps we have going forward is to figure out how do we learn from that kind of informal Sentinel network, the discussions among them, the way that they talk to each other and make sense of what's going on to adapt. And then they often face much more challenging crises than we do in the US in much more vulnerable conditions with much fewer resources. They're somehow able to adapt and show some resilience. And I think there's a lot to learn in how they actually communicate with each other in these kinds of situations. Well, Jared, that's an interesting point. The work that was done in Uganda and the work that was done in the US, it seemed to have, like they had a little bit of a different audience. They were written for or directed at different people. And maybe in the US, there were maybe two audiences at once. Maybe you could talk a little bit about how the audience of these products shapes what's being done. I mean, in both cases, we were trying to inform the US government, FEMA and domestically and USA and internationally about how private sector supply chains were working. But in the US, the audience also, we developed a really rich back and forth with all of the sentinels, these actors who were engaging with us on a regular basis. And so they were as much of the audience as the government. And again, back to the challenge of the fragmented structure in Uganda, we've got to find a way to help the products to develop, get out to the people. They're adapting on their own, but we also need to cultivate those richer networks and be able to share information with them effectively, not just inform where the government can fill gaps and play a role, which is important, but also help them leverage the information that's being gathered so they can, have even more data points to help them adapt and go forward. And in some ways that's in Uganda, they may not be able to see networks across the country. It's very local, the adaptation is very local, but they may not see opportunities with actors in different parts of the country or in different countries as formerly. And that's where again, the work can be more effective if we find a way to serve that audience of informal market actors in those fragmented supply chains in Uganda. I think next I wanna talk a little bit about what it took to do these projects and how maybe the people listening to this podcast, that they were going to try to undertake something similar, what they would need to do to do it. I'll start with you, Tristan, and say, like, what do you think if there was another student who was gonna try to do a project like this, what would they need to kind of make this work? Yeah, I think starting with an understanding of the system is a big advantage and all the work that we had done previously in Uganda was a big help. So I think developing, trying to develop an understanding of the system and basically validating that with system actors is really important. But then in the crisis where things are rapidly changing, I guess having a way to quickly process sort of diverse bits of information has been really important. So developing a methodology for that is really key because, and this is for a crisis specifically because things are changing really fast, right? A lot of incoming information, you want to make the best use of it as you can because there won't be time to do sort of full, rigorous studies a lot of the time. You know, this can come later and can be integrated later, but when things are changing day by day, accepting some uncertainty is really important. And as long as you have a way to sort of deal with that and still make use of the information as best you can. Nice. And maybe I'll turn to you next, Chelsea, to say what Tristan was mentioning, being overwhelmed by events, we often talk about being overtaken by events and trying to have time to synthesize things and make sense of it, but still be able to turn it back to people. Do you have any sort of hints or things you could suggest for people that are trying to do the same sort of work that you did? Yeah, I think that, I mean, we do a lot of preparedness work as well outside of responding to emergencies. And there's a lot of groundwork that needs to be laid prior to an event like this happening or responding to such an event in order to really build out that Sentinel network that you would need. And to Tristan's point, the methodology laying out the methodology prior is really important in being able to adapt as you go as well. So we quickly came up with a cadence that worked for us in order to provide FEMA with the information that they needed. And it allowed us as we went to to figure out how to prioritize the information that was out there. And you've developed a sense for that as you go. So some of it is an adaptation during the process, but I think a lot of it is really happens before the event occurs. Thanks. And then, Jared, maybe I'll leave it to you at the end here to kind of wrap us up and to give us an overall picture here. Well, I think just building on what Chelsea was saying here, there's, you have to do this ahead of time. The preparedness work is important. And this idea, we've talked a lot about systems and supply chains are systems. And maybe it took a health pandemic looking at food supply chains, looking at all the different interconnections of these supply chains to help a broader group of people take an interest in supply chain management and how that's gonna be critical in a crisis like this. And my hope is that then, we can address the crisis quickly and then we can remember to do that preparedness work to formalize what we've learned to build a common vernacular around systems thinking around supply chains and how health supply chains and grocery supply chains and fuel supply chains all connect together. And that, if we have a more common understanding of the structure and especially, a lot of people in the private sector in our supply chain profession have this more innately because they live it. The government may not have it in the same way. And so getting the public and private actors to have a common understanding of those supply chain systems so that then they can anticipate and have a common language to quickly sense make and see what's going on dynamically in a crisis and adapt quicker. That's what we really need to move forward. So I'm hoping that this, it's been remarkable how well the food supply chains have responded to very changing conditions, but we should be ready and do the preparedness work so that we can't adapt better in the future. And we especially having, I don't wanna rest on our laurels that the food supply chain will always be able to be resilient. We may have been fortunate, we should continue studying that and understanding it, but and building that common understanding across public and private sector actors so that they can when faced with a new pathogen or a new global crisis, be able to continue to adapt more quickly to conditions you can't really prepare for, but you can prepare your mindset to be able to react quickly. Well, I thank you for ending on that optimistic note as we head into maybe another COVID spike here in the fall that the grocery supply chain systems are resilient is a nice note to it. And we hope the cold chains that they use will be adapting to get a vaccine out as well. I mean, there's a lot of cross-pollination between these supply chain capabilities. And so some of the actors we worked with on food supply chains are very much central to getting a vaccine out. Well, and these Sentinel networks that you've developed, both in Uganda and in the US, I think will be critical to be able to reach out to them for any needs here in the future. So thanks to everybody for joining and doing this podcast. Really appreciate it. Yeah, thanks, Tim, for coordinating things. Okay, thanks. Thank you, Tim, for the opportunity. All right, everyone. Thank you for listening. I hope you enjoyed this edition of MIT Supply Chain Frontiers. My name is Arthur Grau, Communications Officer for the Center. I invite you to visit anytime at ctl.mit.edu or search for MIT Supply Chain Frontiers on your favorite listening platform. Until next time.