CASE STUDY CONCEPT: Challenges of Disaster Response (DR) Supply Chains and Humanitarian Logistics – decentralized decision-making in unpredictable environments.
You are in charge of coordinating a major disaster response effort. You are working with a geographically dispersed team of people representing different disaster response organizations from government, military, humanitarian and commercial sectors. Your job is to lead people through a collaborative process for designing and operating the supply chains that support the first-responders and people impacted by this flooding disaster.
NOTE: This is an advanced case. Work through the three challenges of the beginning case, “Cincinnati Seasonings” before taking on the challenges in this case.
This case study explores both the expected and unexpected aspects of responding to a regional flooding disaster. First there is the phased deployment in response to predicted and planned for areas of flooding. Then the sudden impending collapse of a dam upstream from a major city brings about the need for rapid response to an unexpected development.
No one organization can create and operate these supply chains entirely on their own, so ways must be found for people from multiple organizations to make decisions and work together. Disaster response missions depend on supply chains to deliver supplies and equipment when and where they are needed. First responders cannot do their jobs without supply chains to deliver the supplies they need. This case study explores creating and managing a disaster response supply chain to handle a flooding disaster similar to the floods that occurred in Central Europe in 2002, 2010 and 2013 (https://en.wikipedia.org/wiki/2013_European_floods).
A Large Scale River Flooding Event is Approaching
It’s been raining heavily off and on for almost two weeks now, and weather forecasts predict even more rain. Water level monitoring sensors on rivers and streams in the region are broadcasting their real-time data showing water levels rising dangerously. As data comes in, the most probable areas of flooding become evident. In planning sessions prior to this date, there have been lively discussions and exchanges of views between the different organizations, so there is a basic plan already in place to respond to this kind of disaster.
Using a cloud-based collaboration platform you convene an online meeting of people from all the organizations that will be part of the disaster response effort. Geographically dispersed teams of people login to the meeting with their PCs and laptops and tablets from wherever they are. You bring up the disaster response supply chain that people have already agreed to (load “Disaster Response Supply Chains – Flooding Scenario” from the SCM Globe library of case studies).
** NOTE: The design for this real-time logistics collaboration platform is presented in our article, “Massively Multiplayer Disaster Response Collaboration“, published by the U.S. Indo-Pacific Command (INDOPACOM) Center for Excellence in Disaster Management, Liaison Journal, Volume 13, Issue 2, Dec 2021. The article starts on pg. 66; view it online or download a PDF copy here – https://drive.google.com/file/d/1K81ShvL7gmvl1XBtVKpPR-1HdOq1VtkQ/view?usp=sharing
It is shown in the screenshot below. All parties can see the same big picture map and the agreed upon supply chain drawn on that map. The supply chain is composed of the four types of entities – products, facilities, vehicles and routes – which have been defined and placed on the map.
Disaster Response Supply Chains Differ from Commercial Supply Chains
Commercial supply chains are based on planning and predictability. Commercial supply chains use continuous data analytics on real-time flows of data to spot trends and predict product demand month by month. This provides the basis for accurate planning and management of supply chain operations. But because disasters are not planned, they are the embodiment of unpredictability. In disaster response (DR) supply chains there is no real-time data to analyze until a disaster actually happens, and nobody knows exactly when a disaster will happen. So DR supply chains cannot always use the same planning and management practices that commercial supply chains use.
DR supply chains focus primarily on responsiveness instead of efficiency. That means timely delivery of needed supplies is more important than doing so at the lowest cost. DR supply chains are usually based on prepositioned stocks of supplies that provide surge capacity for responding to disasters when they happen. DR supply chains call for high levels of problem solving and creative thinking in planning, project management, and logistics in order to respond effectively to the unpredictable nature of disasters.
Attempts to apply centralized command and control in DR supply chains often break down because there are many different organizations involved, ranging from military to governmental, non-governmental, and commercial organizations. These organizations often have slightly different agendas, and they each resist being ordered about by the others. Also, different organizations use different technologies ranging from sophisticated, custom-built hardware and software, to commercial off-the-shelf products such as mobile phones, PCs, and laptops employing email, spreadsheets and social media apps.
Yet all participants must find ways to work together to achieve common goals in order for any DR mission to succeed. This calls for greater levels of human interaction and collaboration, not just more technology. And the technology that is used must be affordable and easily accessible by all organizations involved. This case study illustrates how people, procedures, and technology can be combined to support DR missions. Imagine that all participants are equipped with mobile phones and laptops, and they have internet access so they can all login to the cloud-based collaboration platform described here.
Exploring Options and Reaching Group Consensus
Equipment and supplies have already been prepositioned at distribution centers (DCs) in a proactive move to meet expected flooding scenarios. These DCs are shown in the screenshot below with grey warehouse icons. They are operated by military, governmental, humanitarian and commercial organizations. Transportation of products from these DCs to expected disaster locations (gold store icons) is also provided by a mix of different organizations. Product delivery routes to move supplies from DCs to disaster locations are shown by the blue lines.
First-responders at each disaster location will depend on effective coordination among supply chain organizations to deliver the supplies they need to do their job. The dialog box in the upper right side of the screen below shows information on one of the prepositioned products, “Food Mix” for 100 people. In the sidebar menu on the right side of the screen are edit menus for the four supply chain entities.
[ Instructors, students and professionals can request a free SCM Globe trial demo — NOTE: This is an advanced case. Work through the three online challenges of the beginning case, “Cincinnati Seasonings” before working with this case.]
The working group sees the supply chain model you bring up on your screen. It is already updated with the latest product demand forecasts based on data coming in from the expected flooding locations. After a quick review of this model to orient people, you run simulations so everyone can see how well the planned supply chain the group created earlier will operate in the current actual conditions. The simulation shows day by day performance of the supply chain, and it finds several problems that happen on the fourth day:
- Drinking water in Prague runs out
- Shelter tents in Magdeburg run out
- Drinking water in Riesa runs out
The graph on the right side of the screenshot below shows on-hand inventory at Riesa. The light blue line represents supplies of drinking water. From looking at the map of the supply chain you can see supplies for Riesa are delivered from the Dresden warehouse where they were prepositioned. Simulation data can be downloaded to a spreadsheet for further analysis as shown by the download arrow in the lower left of the screen.
You and the disaster response team use simulations to find and fix problems. You keep updating the supply chain model until you find a solution that works for this current flooding disaster over the next 15 days. Then you fine tune the model to improve its cost and operating performance capabilities as best you can. People quickly reach consensus on this new plan because they were part of creating it, and everyone can see it will work. So that supply chain model that performs well in simulations literally becomes the operating plan for the next two weeks.
The supply chain model shows which facilities will deliver what products to given disaster locations, and it shows the vehicles to be used, and the routes and product delivery amounts and frequencies needed. So each organization then knows what it needs to do based on the facilities and vehicles it operates. Each organizations also understands how their performance supports the group effort and overall success of the disaster response mission.
NOTE: You can explore the situation in Dresden in greater detail to get a more in-depth view, or you can decide to work with this case at the regional supply chain level only. If you decide to work at the regional supply chain level only then skip down to the section heading below titled “FIRST CHALLENGE“.
Dresden Flooding Scenario – Lower Level Detail for a Single City
Let’s take a closer look at operations in and around Dresden to see how local supply chain operations there are progressing. The screenshot below shows a lower level model of the disaster supply chain just in the Dresden area (go to SCM Globe library and import “Dresden Flooding Scenario“). The Dresden warehouse is the same one shown in the larger regional model.
Additional detail is shown for the local aid stations and the vehicles and routes to deliver products from the Dresden warehouse to the aid stations. Notice most of the delivery routes are on roads traveled by trucks. But one delivery route – the straight line between Dresden warehouse and the airport across the river is an air route traveled by helicopters.
People quickly find out more about the aid station across the river by clicking on the Klotzche Aid Station facility. In the screenshot below the dialog box on the left shows information about this facility. The two dialog boxes on the right show information about the vehicles (helicopters) and the route used by those vehicles to deliver supplies to this aid station.
A wealth of data is available to describe the four supply chain entities: products; facilities; vehicles; and routes. Because the data is presented within the organizing context provided by the map-based user interface, this data is quickly understood and people gain situational awareness. Situational awareness allows people to see the big picture and understand how the individual parts of this supply chain work together; they do not get lost in the details.
As you run a simulation of this detailed Dresden disaster response supply chain you see that there is again a problem on day four. The Dresden warehouse runs out of blankets needed to supply the Meissen Aid Station.
All parties in these online sessions can see for themselves how well different disaster response plans work. They can see where the problems are and collaborate on ways to best respond to these problems as they arise in the simulations. Everyone can see what facilities show the greatest demands for products. They can see the vehicles and routes available to deliver those supplies.
This transparency of data, along with the common goal shared by all parties of delivering effective disaster response services, has the effect of bringing about a consensus concerning the actions to be taken by all participants. Different supply chain options are quickly modeled, and simulations show the best courses of action.
Using simulations to explore different options quickly shows everyone what combination of routes and delivery schedules will work best. Because the disaster situation is now understood, people are in a position to fine tune their supply chain operations. Organizations can coordinate with each other to determine who is best situated to deliver what assets to what locations. This drives the coordination between different organizations involved in delivering and using disaster supplies at specific locations.
FIRST CHALLENGE: Create supply chains to deliver supplies from prepositioned DCs to meet demands at flooding locations
Your goal is to design a supply chain that will meet projected demand at all flooding locations for the next 15 – 20 days. And then improve the cost and operating capabilities of this supply chain as much as you can. This will involve trying out different combinations of facilities, vehicles and routes to deliver the products needed at each location.
If you want to add further depth of detail to this exercise, you can create a low level supply chain model that addresses local demands in Dresden. If you need to bring more supplies into the Dresden DC to make your low level supply chain work then be sure to reflect that as higher product demand numbers in the Dresden DC in the high level model. The supply chain must work both at the regional level and the Dresden city level.
As you run simulations of this supply chain watch the graphic and numeric displays for on-hand inventory at different facilities as well as information on products and vehicles that are available by opening the different tabs shown in the display area on the right side of the Simulate screen.
You can take supplies from any distribution center, and use any combination of vehicles and routes. Vehicles can range from trains to trucks and helicopters. If you need more vehicles, go ahead and create them and assign them to facilities where they will be based. Then create the routes they will use to deliver supplies.
In one of the distribution centers (Sankt Polten Distrib Ctr) there is a vehicle called “Ambulance” and that vehicle runs two routes to Linz and Krems. It picks up sick people from aid stations there and brings them back to the DC. DCs have a lot more space than aid stations, so for the 15 or so days of this simulation you can evacuate people from aid stations and concentrate them at the DCs. Assume DCs are set up with beds and supplies for sick and displaced people. If you don’t evacuate sick people from the aid stations, they will build up and aid stations will run out of space (as would happen in a real disaster).
You can create new Ambulance vehicles like the one at the Sankt Polten DC and station them where you think they will work best. You can also make decisions about whether trucks used to deliver supplies to the aid stations can also be used to pick up people from aid stations and return them to the DCs. You can’t put people inside the sealed trailer of a large truck. But some small and medium trucks could be the kind of trucks that have back ends where people can sit. Military trucks are often small and medium trucks that can carry cargo and people. Make some assumptions about how many of your small and medium trucks would be trucks that can pick up people from aid stations and bring them back to a DC.
Here are some things to think about as you design this supply chain:
- Identify the greatest points of risk in the supply chain and modify your design to reduce these risks. Points of risk are facilities, vehicles and routes where a disruption in operations would affect the largest number of other facilities in the supply chain.
- If you are also doing the low level Dresden supply chain and find you need more supplies to make the Dresden supply chain work, then you need to change the high level Central European supply chain so it delivers more supplies to Dresden while still adequately supporting the other facilities in the high level model.
- Unlike commercial supply chains, the best disaster response supply chains are not always the lowest cost. But reducing costs where possible frees up money to be used elsewhere. For useful ideas look in “Reducing Inventory and Operating Costs”.
- THINK OUT OF THE BOX — Experiment with different kinds of vehicles and routes to deliver supplies to the disaster sites. In addition to trucks, railroads and helicopters, consider other types of vehicles such as airships and river barges. What if an airship was available that could travel at 100 km/hr and carry 225 metric tons of cargo? How would that change your supply chain? What if there were places where barges could safely navigate flooded rivers and dock at disaster sites to deliver products and evacuate refugees? What other creative and plausible ideas can you come up with?
- We live in a new 21st century world where unexpected things happen more frequently, so thinking out of the box isn’t just a tired old phrase… it’s a critical skill for disaster response.
When you have questions about how to work with this case, the answer is always to ask yourself, “What would I do if this were the real world and I was the person in charge?” Then you model and simulate different ideas. That means you make plausible assumptions, then add or change facilities, vehicles and routes as called for to model your ideas. Simulations will show how well they work. Pick the best one.
CREATE A SHORT EXECUTIVE BRIEFING — a 3 to 5 page report or a short deck of presentation slides. Use screenshots and data produced by simulations to illustrate what you learned. Explain what were the main problems you faced in getting your simulation to run for 15+ days. Show what you did to fix those problems. And present the three or four main things you discovered about this how this supply chain works. What ideas do you have for improving the performance of a disaster response supply chain such as this?
MISSION REPORTING TEMPLATE — There is a reporting template with an operations report and a performance dashboard for analyzing your simulation data. The template is built for a 15-day period. You can import your simulation data after trimming it for 15 days (scroll down to bottom of Analyzing Simulation Data). The operations report shows facility and product detail, and the dashboard shows where the best opportunities are for improvement. You can download a copy of the mission reporting template here.
The reporting template is set up for the Nepal Earthquake disaster response supply chain, but look at how the reports read the product and simulation data – you will see how to change the spreadsheet as needed to accommodate this case study.
NOTES FOR USING THE MISSION REPORTING TEMPLATE — In addition to the original P&L report and KPIs, there are new features and tabs in the Mission Reporting template as described below:
- Ordering cost – an estimate of how much it costs an organization to place an order, receive the order, issue payment and put away the inventory delivered. Different companies have different ordering costs; estimates can range from $15 to $150 or more. Industry trade associations can provide data on average ordering costs for companies in their industries, or your company may already have an ordering cost it uses for EOQ calculations. The template uses a default cost of $35.00.
- Annual holding costs – often estimated at 20% of product price. Industry trade associations can provide data on average annual holding costs for companies in their industries.
- EOQ Delivery Amount and Frequency — these features use data from the supply chain model plus the ordering cost and holding costs to apply the Economic Order Quantity (EOQ) equation to suggest an amount and frequency for product deliveries to different facilities. The numbers are only suggestions and often need to be adjusted, but they provide a starting point. See more about this in the online guide section “Cutting Inventory and Operating Costs“, scroll down to the heading “Use Economic Order Quantity to Calculate Delivery Amounts and Frequencies”
- Dashboard Tab — a high level view of overall supply chain performance. You assign weights to each of the performance categories (Inventory, Transportation, and Storage) to reflect your strategy for improving operations. Assign higher weights to the categories most critical to the success of a particular mission. This will focus people’s attention on the performance categories you most want to improve. Assign category performance targets that are ambitious but achievable. The dashboard calculates percent effectiveness for each facility and the overall plan by comparing the simulation results generated by your supply chain model to your predefined performance targets. Based on the weights assigned to each target, a net performance score is calculated for the whole supply chain plan. Performance targets are recommended by logistics planners and approved by the mission commander.
- Products Tab — enter the price, weight, and volume for each product in the supply chain model. These numbers will be used by the macro equations in the first two tabs to calculate supply chain KPIs and Dashboard displays.
[If you are using SCM Globe Professional version, mission reports can be generated automatically by clicking on the “Generate P&L Report” button on the Simulate Screen]
SAVE BACKUP COPIES of your supply chain model from time to time as you make changes. There is no “undo”, but if a change doesn’t work out, you can restore from a saved copy. And sometimes supply chain model files (json files) become damaged and no longer work, so you want backup copies of your supply chain to restore from when that happens.
To share your changes and improvements to this model (json file) with other SCM Globe users see “Download and Share Supply Chain Models”
CASE CREDIT: This case is based on work done by Dr. Dennis Duke, Florida Institute of Technology, and Michael Hugos, SCM Globe, in a paper they presented at the ITEC Conference in Prague, April 2015 titled “Supply Chain Simulations for Training Disaster Response Managers“. You can download a PDF of the paper here
More resources for exploring disaster response supply chains are available on the SCM Globe website:
- See model and simulation of the supply chain used by the World Food Program mission for the Nepal Earthquake disaster in 2015 – Nepal Earthquake Disaster Response – 2015
- Examine models and simulations of humanitarian and military supply chains to support a massive humanitarian intervention in a Syrian peace settlement – Humanitarian Supply Chain – Syria Evacuation (CIV and MIL Supply Chains)
- For additional information on modeling techniques you can use in this case study see “Tips for Building Supply Chain Models”
SECOND CHALLENGE: An Unexpected Event calls for Surge Capacity
You put the supply chain model created above into operation. And disaster response managers from different organizations are able to respond effectively to demands in the expected flooding areas. The disaster situation seems to be in hand; good progress is being made delivering the supplies needed at each of the flooding sites. It looks like you have the next 15 days covered.
A few days go by, and you start to relax a little. But suddenly something unexpected happens — as it so often does in disaster situations. After weeks of rain the main dam on the Vitava River upstream from Prague is showing signs of distress. The hillsides on either side of the dam are sodden with water and there seem to be cracks appearing in the wall of the dam on one side of the river. If the dam were to give way a 7 meter (21 foot) wall of water would go rushing down the river valley and reach Prague in about half an hour.
Reports keep coming in hour by hour that the situation at the dam is deteriorating. There are perhaps two days, maybe three before the engineers expect the dam to give way. To avert a complete collapse, the flood gates must be opened in the next 72 hours to relieve the pressure on the dam. But that will still create a surge of water only a little less destructive than the collapse of the dam itself. This creates a sudden and urgent demand for supplies to be delivered to Prague in the next 72 hours.
Your challenge is to modify the high level supply chain model you created for the first challenge so as to respond to this new development. You still have to support the existing operations as shown in your model, and now you also have to respond to this new need for supplies in Prague. You click on the Prague Warehouse facility in the sidebar menu on the right of the screen to get a display of on-hand inventory and other information. It is clear that supplies in Prague Warehouse have already been committed to cover demand in Pisek, Prague, Usti nad Labem, and Decin.
Find Additional Supplies and Equipment
You can see from the situation map that the distribution centers in Frankfurt, Munich, Sankt Polten, Dresden and Berlin are already committed to supporting their respective disaster sites. New supplies must now come from currently uncommitted warehouses in Wroclaw, Krakow, Bratislava and Budapest. People have to balance the need to provide aid to a neighboring city, with the need to maintain enough supplies on-hand to respond to a disaster that might befall their own cities.
Create a supply chain to take supplies from available distribution centers and deliver needed quantities to the Prague Warehouse within 72 hours. Simulate its operation to show that it will perform as needed to meet this challenge.
To investigate this extension to the supply chain you can do a quick simulation by adding new products, vehicles and routes to the existing supply chain model. You can run a simulation that will show two days of data and show if the needed products arrive in Prague within 72 hours. Or you can adjust the timeline to explore this 72 hour event in more detail.
SURGE CAPACITY MODEL – Detailed Hourly Data for a Single Event
In your supply chain model you can shrink the timeline and make one day equal a half day or a quarter day or less. To do this, import a new copy of “Disaster Response Supply Chains – Flooding Scenario” from the SCM Globe library. Remove all facilities except the facilities in Prague, Wroclaw, Krakow, Bratislava and Budapest. Now add new products as described below, and add new vehicles and delivery routes.
Edit attributes of the facilities (daily demands and costs) and vehicles (speed and delay between departure) to set the timeline for the simulation at 6 hour intervals, not the standard 24 hour intervals normally used by the simulations. Now the simulations will provide more detailed hourly data which is appropriate when hours count as they do in this second challenge. Instructions for how to adjust the timeline of a simulation are in “Tips for Building Supply Chain Models” — scroll down to the heading “SIMULATION TIMELINE”
There is need for flood barricades (1000), sandbags (1,000,000), bulldozers (24) and cranes (12). And the need for food and shelter for people who will have to be evacuated from areas of Prague causes existing demand numbers in Prague to triple. Here are some questions to explore as you simulate possible solutions to this challenge:
- What distribution centers have how much of these existing supplies and how much can be delivered from those warehouses in the next 72 hours?
- Add new products to this supply chain model: sandbags, bulldozers, cranes, and watertight barricades. Do some quick research and use your best estimates to define their cost, shipping weight and size, and position them at the four uncommitted warehouses. Assume each of these four warehouses has enough on-hand inventory of these new products to meet the needs in Prague.
- Then create new vehicles and routes to bring the new products from their points of origin to Prague. What additional types and numbers of vehicles are needed?
- Which or what mix of the available uncommitted warehouses should respond to this need for surge capacity?
- What are the risks of other unexpected developments that would call for these same warehouses to respond in their own cities?
- From your experience with simulations of different scenarios in this case study, how much extra capacity (surge capacity) do you think should be built into the on-hand inventory levels at the distribution center warehouses in this disaster response supply chain to handle unexpected events in the future? Why?
CREATE A SHORT EXECUTIVE BRIEFING — a 3 to 5 page report or a short deck of presentation slides. Use screenshots and data produced by simulations to illustrate what you learned. Explain your answers to the bullet point questions above.
A Disaster Response Collaboration Platform Accessible by All
This case study illustrates use of off-the-shelf technology to create an online collaboration platform that all parties can access and use. It uses simulations to drive supply chain planning, and provide training for organizations to learn to work together in disaster response situations.
The design for this real-time logistics collaboration platform is presented in our article, “Massively Multiplayer Disaster Response Collaboration“, published by the U.S. Indo-Pacific Command (INDOPACOM) Center for Excellence in Disaster Management, Liaison Journal, Volume 13, Issue 2, Dec 2021. The article starts on pg. 66; view it online or download a PDF copy here – https://drive.google.com/file/d/1K81ShvL7gmvl1XBtVKpPR-1HdOq1VtkQ/view?usp=sharing
Ideas presented in this case are developed further in another disaster response case study called – Nepal Earthquake Disaster Response Supply Chain. The case is based on work with the Global Logistics Cluster of the World Food Program. It simulates the supply chain created by the World Food Program to support their humanitarian mission to the Nepal Earthquake disaster in the spring of 2015.
Register on SCM Globe to gain access to this and other supply chain simulations. Click the blue “Register” button on the app login page, and buy an account with a credit card (unless you already have an account). Scan the “Getting Started” section, and you are ready to start. Go to the SCM Globe library and click the “Import” button next to this or any other supply chain model.