modeling disruption in a fresh produce supply chainmodeling disruption in a fresh produce supply...

Department of Industrial and Manufacturing Systems Engineering

Modeling Disruption in a Fresh

Produce Supply ChainCameron MacKenzie, Assistant Professor, Industrial and Manufacturing

Systems Engineering, Iowa State University

Aruna Apte, Associate Professor, Graduate School of Business and

Public Policy, Naval Postgraduate School

POMS Annual Conference

May 6, 2016

Industrial and Manufacturing Systems Engineering

Fresh produce contamination

2


Previous research

• Uniqueness of fresh produce supply chains

• Factors that influence fresh produce supply chain

disruptions

• Traceability, transparency, testability, time, trust, and

training (Roth et al., 2008)

• Traceability, product type, communication, topological

structure, exposure to contamination (Apte, 2010)

• Disruption management mathematical models

3

Roth, A. V., Tsay, A. A., Pullman, M. E., and Gray, J. V. (2008). Unraveling the food

supply chain: Strategic insights from China and the 2007 recalls. Journal of Supply

Chain Management, 44(1), 22-39.

Apte, A. (2010). Supply chain networks for perishable and essential commodities:

Design and vulnerabilities. Journal of Operations and Supply Chain Management,

3(2), 26-43.


What’s new with this research?

• Build a model that quantifies the effects of a

contamination in fresh produce supply chain

• Translate qualitative factors that influence

vulnerability into a mathematical model

• Quantify benefits of disruption management

strategies for contamination of fresh produce

• Apply model to Dole’s supply chain during 2006

E. coli contamination

4


Model flowchart

5

At demand

node

Where is

contamination

node (CN)?

CN is distribution

node

CN is producing

node

Can produce

be rerouted?

Is safety stock

useable?

Is safety stock

useable?

Can

production be

increased?

Disruption

mitigated

Disruption

not

mitigated

Disruption

mitigated

Disruption

not

mitigated

Yes

Yes

No

Yes

Yes

No


Measure of impact

• Lost sales compared to sales without

contamination

• Sum of three components

1. Supply chain closed which searching for

source of contamination (traceability)

2. While relying on safety stock (perishability

of safety stock, rerouting production,

customer demand)

3. After safety stock depleted (rerouting

production, increase production, customer

demand) 6


Traceability and prior probability

7


Topology of supply chain

8


Safety stock and demand

Contaminated node is

distribution node

Contaminated node is

producing node

9

Production before disruption Production before disruption


E. coli in bagged spinach

• E. coli discovered in bagged spinach in

September 2006

• Source of contamination traced back to Natural

Selection Foods, a supplier to Dole

• Fresh and bagged spinach was pulled from

shelves for 5 days

• Spinach from California unavailable for an

additional 10 days

10


Dole’s production of spinach

11


Conclusions

• Model different elements in a supply chain that

affect vulnerability

• Traceability: length of time supply chain is closed

• Communication and essentialness of product: demand

during disruption

• Topology: ability of supply chain to reroute production

• Can help supply chain managers quantify benefits of

different disruption management strategies

• Optimal value of safety stock

• Communication to keep demand high

• Useful for risk management: cost and uncertainty

12Email: [email protected]


Backup

13


Traceability of contamination

• Assume entire supply chain is closed while

finding source of contamination

𝑇1𝑄

• Traceability may depend on

• High-tech growers

• Supply chain visibility

• Prior probability of node being contaminated

14

Time to find

source of

contamination

Amount

produced before

contamination


Demand after contamination

• Demand for product may drop because of

contamination

• New demand after contamination 𝐷∗

• Depends on

• Essentialness of product

• Communication to consumers

• Trust level

15


Safety stock

Safety stock may be used to mitigate disruption

min 𝑡𝑠 − 𝑇1+,

𝑠

𝐷∗ − 𝑄∗, 𝑇2 𝑄 − 𝐷∗

16

Time to find

source of

contamination

Time until

safety stock

perishes

Amount of safety stock

Amount

delivered after

contamination

Amount produced

before

contamination

Demand after

contaminationTime until

contaminated node

reopens


After safety stock depleted

If contaminated node is distribution node

𝑇2 −min 𝑡𝑠 − 𝑇1+,

𝑠

𝐷∗ − 𝑄∗

+

𝑄 − 𝑄∗

17

Time to find

source of

contamination

Amount

of safety

stock

Amount produced

before

contamination

Time until safety stock perishes Amount delivered

after

contamination

Demand after

contamination

Time until

contaminated node

reopens


After safety stock depleted

If contaminated node is producing node

𝑇2 −min 𝑡𝑠 − 𝑇1+,

𝑠

𝐷∗ − 𝑄∗

+

𝑄 − 𝑄∗∗

18

Time to find

source of

contamination

Amount

of safety

stock

Amount produced

before

contamination

Time until

safety stock

perishes

Amount delivered

after contamination

without additional

production

Demand after

contamination

Time until

contaminated node

reopens

Amount delivered

after

contamination with

additional

production

modeling disruption in a fresh produce supply chainmodeling disruption in a fresh produce supply...

Documents