Strategic and Standardized Simulation of a Distribution Network

A Case for a Drugstore Company in Mexico

Homero H. Contreras

, José Pablo Nuño

, Eric Porras

and Eduardo Zelaya

Centro Interdisciplinario de Posgrados, UPAEP University, Puebla, Pue., Mexico

EGADE Business School, ITESM-Santa Fe, México, D.F., Mexico

Keywords: Simulation, Standardized Model, Multiechelon Inventory, Distribution Network.

Abstract: Analysis of distribution network is a crucial issue in supply chain management. There is a vast array of

analysis tools for logistics, but analytic tools cannot deal with the inherited variability. Thus, simulation

might be a better alternative, and the use of standardized models represents a promising areas. In this paper,

a simulation model facing a strategic approach will be proposed as a way to analyze a distribution network

based on model consisting of two-echelons; this model can work both forwards and backwards in a

recursive manner, and relies on operative key performance indicators that affect the strategy in the long

term. Using a standardized model increases flexibility, focus the problem and provides a better computer

performance. The model is validated through a business case for a Mexican company dealing with bottom

of pyramid clients in the drugstore sector.

1 INTRODUCTION

Supply chain management is a challenge for every

organization, and in the case of healthcare, suitable

supply of medicines is a crucial issue, existing

opportunities to improve distribution under a

strategic approach using simulation techniques.

This paper is organized as follows: first, a

literature review of strategy and simulation related to

supply chain is presented; then, the focus of a

standardized simulation model for distribution

networks in a drugstore company is presented,

followed by a test of this model in a real case in

Mexico and finally, some conclusions and further

research are mentioned.

2 LITERATURE REVIEW

2.1 Strategy

Strategic planning is a key issue for companies in

today’s world competition. It provides a guide to

achieve sustainable results, and all activities within

the companies must be aligned to their strategy, as

stated by Porter (1996). Then, supply chain

management must support the strategy of the

company, but it should be considered as “a dynamic,

stochastic and complex system” (Pundoor and

Hermann, 2006).

2.2 Supply Chain

One strategic objective of supply chain is “fulfilling

customer demand, assuring the on-time delivery of

high-quality products at a minimal cost and with the

minimum lead-time” (Chang and Makatsoris, 2001).

In the healthcare industry, having the required

products at any moment of time is vital and also

represents a social activity in the community served.

Hung, Kuchereko, Samsatti and Sha, 2004, have

suggested that supply chain must deal with external

strategic challenges and also with the operational

uncertainty, so new opportunities can be detected.

2.3 Simulation and Supply Chain

Supply chain must be evaluated in an effective

manner, and simulation can be used within the

supply chain to analyze different issues: inventory

policies, configuration of activities, distribution, lead

times, costs, etc.

Shanthikumar and Gargent (1983) claim that

simulation models are focused on dynamic models

that resemble the behavior of a real system; and

445

H. Contrera H., Pablo Nuño J., Porras E. and Zelaya E..

Strategic and Standardized Simulation of a Distribution Network - A Case for a Drugstore Company in Mexico.

DOI: 10.5220/0003997604450448

In Proceedings of the 2nd International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH-2012),

pages 445-448

ISBN: 978-989-8565-20-4

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

Buxton, Fuqua and Wyland (2000) that supply chain

must be flexible enough to adapt to a wide range of

potential futures. Petrovic, Roy and Petrovic (1998)

even consider a crucial issue the separation of

strategic issues from those tactical or operational

ones.

2.4 Standardized Models

Pundoor and Herrmann (2006) suggest that, no

matter the components of the supply chain, there are

always some common processes that can be reused,

leading to the definition of standardized simulation

models. Jain, Collins, Workman and Ervin (2001)

suggest the use of generic tools to define a flexible

model of the supply chain. Cope, Sam-Fayez,

Mollaghasemi and Kaylani (2007) emphasize on

using generic simulation models that can be

reconfigured in an easy way to individual projects,

while Brown and Powers (2000) suggest simulation

models focused on a specific problem, but flexible

enough to evolve in the future, as well as Longo and

Mirabelli (2008) and Petrovic, Roy and Petrovic

(1998).

3 SIMULATION MODEL

Considering a supply chain for a drugstore company,

we propose a standardized simulation based on a

two-echelon structure that can be reused on other

parts of the chain, thus allowing a complete analysis,

and under a strategic approach, to support relevant

decisions in the long term. The assumptions of the

model are:

3.1 Standardized and Recursive Model

A set of common operations has been defined and

will be used as the logic to be standardized and

recursively used both forwards and backwards.

3.2 Strategic Approach

The model must encompass the long-term focus of

the distribution network, but also includes two

tactical or operational indicators in order to evaluate

the network, being:

1. The location of inventories within the

network (and their associated levels).

2. The transportation cost.

The integration of both indicators will provide a

total cost, consisting of total inventory holding cost,

transportation cost and the financial cost of

inventory.

The strategic consideration is supported by the

use of an aggregate demand.

3.3 Unitary Transportation Cost

A unitary transportation cost, considering the routes,

vehicles and aggregated amount of products

transported, will be calculated.

3.4 Discrete Operation

All variables are transformed to observation based

one, and the model is also based in a non-temporal

time framework. The model is based on a single

control entity that flows through the model and

executes each of the logic steps defined.

3.5 Standard Logic

The model encompasses some common processes

found during the inventory and replenishment

systems in all major SC systems. The main logic of

the model is embedded in a two-echelon framework,

including non-strategic operations which affect the

strategic deployment, and will be based in a one

week time period.

The two-echelon logic can be replicated to a

series of clients-suppliers in different parts of the

supply chain, where a supplier becomes a client of

another supplier. The replication, both forwards and

backwards, can also be replicated in a parallel

framework. The code is generic in a sense that is

programmed only once for the common logic, and

then can be reused, as can be seen in Figure 1.

Figure 1: Standard model and recursion.

3.6 Simulation Software

The previous logic might be so complex to be

performed by a graphical simulator, so it was

developed in the simulation language SIMNET II,

owed to Dr. Hamdy Taha, which provides the so-

called PROCEDURES that can be considered as a

standard part of the code that can be automatically

SIMULTECH 2012 - 2nd International Conference on Simulation and Modeling Methodologies, Technologies and

Applications

446

replicated both in series or parallel, providing a

generic and reused code.

4 TESTING THE MODEL

4.1 Validation

A business case was developed using a drugstore

company in Mexico, focused to serve the Bottom of

Pyramid population, in order to test and validate the

model based on its actual distribution network.

This company manages its distribution network

through a master distribution center (MDC), which

serves nine regional distribution centers (RDC).

Each RDC serves specific regions. The base unit of

aggregation of demand will be a region.

Furthermore, any region is formed by local

warehouses and stores (some owned by the company

and other ones are franchisees). There are a total of

35 warehouses and about 4,000 stores located across

Mexico. A brief schema of the distribution network

is shown in Figure 2:

Figure 2: Layout of distribution network.

At any region, the RDC serves owned stores, big

franchisees and local warehouses. There is no

transportation cost to small franchisees, because they

must pick up their products.

The MDC is used as a delivery station for

products manufactured by the company in its private

laboratory, located next to the MDC.

Transportation between MDC and RDC, and also

between RDC and servicing facilities are carried out

by an external company.

Each RDC operates independently from the

others; there is no overlap in and therefore a two

echelon network divided in two phases can be

considered:

Phase 1: Each RDC and its associated region.

Phase 2: The MDC and its associated RDC.

The model can be used as nine-independent analysis

of the RDC and regions, where the RDC is the

supplier and each region is a client. Then, the RDC

become clients of the MDC in another analysis, as

can be seen in Figure 3:

Figure 3: Recursive use of the standardized model.

Data from one complete year was available and

used for validation purposes. Once the regions were

defined, model was tested under the actual policy of

30 day of stock in inventory levels and under

steady-state analysis, and the differences in the total

inventory level were about 2.7% versus historical

data. Figure 4 presents comparisons versus

simulation and historical data based on a percentage

basis, where history is represented by 100%.

Figure 4: Comparison of inventory level of base case

versus simulation model.

In the case of transportation cost, difference of

simulation versus historical data is about 3.2%, as

can be seen in Figure 5:

Figure 5: Comparison of transportation cost of base case

versus simulation model.

Strategic and Standardized Simulation of a Distribution Network - A Case for a Drugstore Company in Mexico

447

Considering this differences and a target error of

5%, results from simulation model is within

tolerances.

4.2 Improvement Case

The model was used to analyze some alternatives to

improve the distribution network, including:

 Opening/Closing/Merging of RDC.

 Reassignment of regions to RDC.

 Adding delivery frequencies.

 Using multiechelon inventories.

About twenty scenarios were simulated and the

model only required adding a small code to manage

the multiechelon inventory and the additional

delivery frequencies. A new distribution network

was found, composed by eight RDC, using

multiechelon inventories and serving twice-per week

to the metro areas, as shown in Figure 6:

Figure 6: Configuration of proposed distribution network

after simulation analysis.

Considering the actual distribution network as

100%, there are significant savings, as shown in

table 1:

Table 1: Comparison of key performance indicators of

base case versus proposed network.

The final savings of 19.3% of the total costs is

important for the company; this savings can be used

to reinforce the competitive position of the firm.

5 CONCLUSIONS

The proposed model, based on a two-echelon system

that can be replicated both forwards and backwards,

has been tested and used in a real situation to

improve a distribution network; its operation has

been fast, and helped to focus on the most important

characteristics of the model.

Finally, it is really critical to consider that any

simulation model, specially a strategic one, must be

designed to support a company to comply with its

strategy. This model must fit within the “reducing

cost” strategy to improve service to the bottom of

pyramid clients.

REFERENCES

Brown, N. and Powers, S. 2000. “Simulation in a box (a

generic reusable maintenance model)” Proceedings of

the 2000 Winter Simulation Conference, Edited by J.

A. Joines, R. R. Barton, K. Kang, and P. A. Fishwick,

1050-1056. Piscataway, New Jersey: Institute of

Electrical and Electronics Engineers, Inc.

Buxton, K., Fuqua, B. y Wyland, B. 2000. “Simulating the

supply chain”. IIE Solutions, 32(1):37-42

Chang, Y. y Makatsoris, H. 2001. “Supply chain modeling

using simulation”. International journal of simulation

systems, science and technology, 2(1):24-30

Cope, D., Sam-Fayez, M., Mollaghasemi, M. And

Kaylani, A. 2007. “Supply chain simulation modeling

made easy: an innovative approach”. Proceedings of

the 2007 winter simulation conference, Edited by S. G.

Henderson, B. Biller, M.-H. Hsieh, J. Shortle, J. D.

Tew, and R. R. Barton, 1887-1896

Hung, W Y., Kucherenko, S., Samsatli, N. y Shah, N.

2004. “A flexible and generic approach to dynamic

modelling of supply chain”. Journal of the operational

research society, 55:801-813

Longo, F., and Mirabelli, G. 2008. “An advanced supply

chain management tool based on modeling and

simulation.” Computers & Industrial Engineering,

54:570-588.

Petrovic, D., Roy, R., y Petrovic, R. 1998. “Modelling and

simulation of a supply chain in an uncertain

environment”. European journal of operational

research, 109:299-309

Porter, M. (1996). “What is strategy?” Harvard Business

Review, 74(6):61-78.

Pundoor, G. And Herrmann, J. 2006. “A hierarchical

approach to supply chain simulation modeling using

the supply chain operations reference model”.

International journal of simulation and process

modelling. 2(3-4): 124-132.

Shanthikumar, J. y Gargent, R. 1983. “A unifying view of

hybrid simulation/analitic models and modeling”.

Operations research, 31(6):1030-1052

Indicator Actual Proposed Savings

Inventory (pieces.) 100% 50.1% 49.9%

Transportation cost $$$ 100% 98.6% 1.4%

Total cost $$$ 100% 80.7% 19.3%

SIMULTECH 2012 - 2nd International Conference on Simulation and Modeling Methodologies, Technologies and

Applications

448