1. A Supply Network Resiliency Assessment Framework
Authors: Jaspar Siu, Santosh Stephen
Advisor: James B. Rice, Jr.
MIT SCM Research FEST
May 21, 2015

2. Supply Chain Resiliency
- the ability of a supply chain network to bounce back from a disruption.

3. Why is Supply Chain Resiliency Important?
Albuquerque, NM
Impact:
Production down for weeks
$M’s of chips destroyed
Sole Sourced Supplier
Phone Manufacturer
Response
Fast, Proactive
Slow, Reactive
Business Impact $
Market share rose
from 27% to 30%
Market share fell from
12% to 9%, lost $2.34 Billion
Source: Yossi Sheffi, “The Resilient Enterprise”, MIT Press, 2005

4. Agenda
Problem
Approach
Visualization
Results & Insights

5. Problem with Resiliency today
Based on gut-feel and intuition
Lack formal process to evaluate and visualize supply chain risk resiliency
Unclear where to spend mitigation dollars and how much
Problem
Approach
Visualize
Results

6. Quantify resilience in supply chain
3 Key Takeaways 1
Quantify resilience in supply chain
Visualize supply chain risk
Quantify the value of mitigation options 2 3
Problem
Approach
Visualize
Results

7. Agenda
Problem
Approach
Visualization
Results & Insights

8. Definitions: Timeline of a Disruptive Event
TTR: Time to Recover
7 days
20 days
30 days
13 days
10 days
TTB: Time to Backup
Timeline
Steady State
Steady State
Disruptive Event
Disruption
Backup Supplier
Baseline
Supply
Impact
Downstream Inventory
Black out period
Back up period
Before we understand the mechanics of how we quantitatively assess resiliency, let me first level-set with a timeline and some definitions.
This is a stylized timeline for a disruption and subsequent recovery at an upstream supply chain.
Below that is the corresponding business impact to the company.
- The supply chain is in steady state before a disruptive event, such as a fire or an earthquake, disrupts the productive capacity at a node.
- At this point, even though the node has experienced a disruption, we are still able to meet downstream customer demand from inventory.
- When the inventory runs out, the demand chain is disrupted, and we enter the BLACKOUT PERIOD, where we are unable to meet customer demand anymore
- At THIS point, after TTB days, the back up supplier comes online, and the customer demand is met. However, during this Backup Period, it typically costs more to operate than the steady state before disruption. And at THIS point, after TTR days, either the disrupted node or the alternate supplier comes fully online and operates at normal cost, and we say that steady state operations have resumed. During the BACKUP PERIOD, the
The overall Business Impact is the sum of the loss of potential sales contribution during the Blackout Period and the increased cost of operation during the backup period.
We understand that in different companies, the disruptive process and the recovery process can be different – but this is a general view that can be customized to your specific situation.
To help you understand this better, consider a simple example:
If there are 7 days of inventory, and the time to backup is 20 days, and the time to recover is 30 days, then the disruption occurs at 7 days, and the BOP is for 13 days, and the BUP is for 10 days.
Business Impact = Lost Sales Contribution + Increased Cost
Problem
Approach
Visualize
Results

9. Explanation: Computing Expected Business Impact
Total
Business
Impact
Probability
Geo-political risk
Lost Contribution
Cost Increase
Natural Disaster risk
Expected Business Impact is the product of the Total Business Impact times the probability of the disruptive event.
The total business impact, as we saw before, is the sum of POTENTIAL LOST CONTRIBUTION and the increased cost of operation. We compute this for each type of risk event and multiply by its corresponding probability of occurrence to get the total expected value of business impact across all events.
Supplier risk
Blackout Period X
Part Volume Rate
Contribution
/ Unit
Backup Period X
WIP Volume Rate
Cost Increase
/ Unit
Process risk
Problem
Approach
Visualize
Results

10. Example: Expected Business Impact at a Node
Each node in our supply chain is a supplier-facility-process. We carry out the expected BI computation at each node in the supply chain, and also aggregate it across multiple sets of nodes such as entire suppliers or locations.
We will now look at an example of the expected BI computation as a single node.
Problem
Approach
Results
Insights

11. Example: Expected Business Impact at a Node
Process failures
Minor quality issues
Example
Minor
Major
Relatively low impact
Medium or high probability events
Description
Earthquakes
Vendor bankruptcy
High impact
Low probability events
For the sake of simplicity, we combined and categorized the entire set of risk events into minor and major events.
- Minor events are events such as process failure or minor quality issues that have lower Time to Backup and Time to Recover, and therefore lower impacts, but they have greater probability of occurrence.
- Major events are events such earthquakes or vendor bankruptcies that have higher Time to Backup and Time to Recover, and therefore higher impacts, but they have rarer events.
This table shows a simple Expected BI computation for a single node.
For this node, for minor disruptive events, the TTR is around 20 days, and the probability is around 40%. This gives an expected business impact of $1500 for minor events.
A similar computation fives the expected BI for major disruptive events as $750. Therefore the total expected BI for all major and minor events is the sum of $1500 and $750, which is $2250.
Important to node that for this node, the expected BI from minor events is greater than the expected BI from major events. This is because even though a single minor event has lesser impact than a single major event, minor events are much more common and frequent than major events.
Problem
Approach
Visualize
Results

12. Agenda
Problem
Approach
Visualization
Results & Insights

13. ABC Company’s Supply Chain – One Raw Material Commodity
Facts:
Used in 134 parts
3 suppliers
6 locations
3 stages of processing then ODM
Problem
Approach
Visualize
Results

14. Map view: Expected Business Impact Risk
Sourcemap enables supply chain visualization for:
Node locations
Directional flow of material
Business impact metric
Color-coded
Problem
Approach
Visualize
Results

15. SC network view: Expected Business Impact Risk
SC Network Visualization allows us to see:
ODM ~ 1.73M exp. BI
Supplier 1 highest exp. BI
Assembly process highest exp. BI
Color code – risk levels at a glance
ODM ~ 1.73 million exp. BI risk {Talk about nodes, not aggregates}
- Downstream nodes have higher BI, since the total downstream inventory is lesser. As we go upstream, the total downstream inventory increases and reduced the BOP, which in turn reduces the expected Contribution at Risk: Supplier 1A, 1B
- Nodes that have higher volumes flowing through them have higher expected BI: ODM, Supplier 2 Assembly
- Supplier 2 Assembly has lesser volume flowing through it, yet has higher exp. BI due to lesser inventory at location. This is because both that node and downstream ODM node are co-located
Problem
Approach
Visualize
Results

16. Agenda
Problem
Approach
Visualization
Results & Insights

17. Results: Identify and Prioritize Risky Locations and Suppliers
Risk by facility ($K) A B D
ODM 1A 1B
2–R2
2-R1
3-R1 F
Identifying Critical Entities:
Supplier 1
Location A
Supplier 1: ~ $600k due to 2 facilities
Location A: ~ $2M due to concentration of suppliers
Risk by location ($K) A B C D E F
Risk by supplier ($K)
ODM 1 2 3
Problem
Approach
Visualize
Results

18. Results: Quantifying Value of Mitigation Options
$200K decrease per day
$10K decrease per day
Explain why up to 7?
Problem
Approach
Visualize
Results

19. Results: Quantifying Value of Mitigation Options
$1K decrease per day
$12K decrease per day
Explain why up to 30 days?
Problem
Approach
Visualize
Results

20. Insights Tension is present between efficiency and risk.
Visualizing supply chain risk helps managers understand geographic location risk. Risk aggregates when same suppliers or locations occur multiple times.
Risk depends on Time-To-Recovery (TTR), Time-To-Backup (TTB), downstream inventory, supply chain structure, and volume of flow.
Choice of mitigation option, and extent of investment depends on marginal benefit of option vs. additional cost (illustrated from response curves).
Tension is present between efficiency and risk. The leaner and more “procurement cost optimized” the supply chain is, without providing for back-up plans, the less resilient it may become.
Visualizing supply chain risk helps managers understand proximity of risk. Risk aggregates when same suppliers or locations occur multiple times.
Risk at a node depends on Time-To-Recovery (TTR), Time-To-Blackout (TTB), downstream inventory, and supply chain structure, and volume of flow.
The choice of mitigation option, and extent of investment depends on marginal benefit of option vs. additional cost incurred from response curves.
Problem
Approach
Visualize
Results

21. 3 Key Takeaways 1
Quantify resilience in supply chain
Visualize supply chain risk
Quantify the value of mitigation options 2 3