Understanding of Logistics Network Optimization - Why and How to Optimize a Logistics Network?

1. Definition & Scope of Logistics Network Optimization

Logistics network optimization is about determining the number, location and size of warehouses that are optimal for each business by taking into account a wide range of constraints in one’s supply chain. In other words, it is to find the best combination of warehouses necessary to cover the entire supply chain from raw material suppliers to end-users.

Logistics takes place largely in three areas - procurement, supply and sales. Typically, logistics network optimization is for supply and sales of finished goods, although it sometimes includes sourcing of parts or raw materials. In the end, logistics network optimization aims to find an optimal combination of various warehouses best suited to the client’s needs.

Of course, manufacturing factories can be part of a logistics network since they do have in-house warehouses and temporary storage. However, given that when building a factory, the main priorities are usually on terms & conditions of the site acquisition, tax benefits, political risks, market size, or personnel sourcing while logistics factors are considered next, where to locate factories is relatively less important in logistics network optimization.

Why It Needs & What to Expect

Network optimization is a top strategic project in logistics. It is also a critical issue that requires decision-making by senior leadership because it affects a company’s investment in infrastructure and entire supply chain. In the below, you can find major reasons why network optimization is needed. .

- The addition of new sales channels in a country
- Increases/decreases in demand (by product and/or geography)
- Changes in manufacturing capabilities (relocation/opening/closing of factories)
- Changes in retail channels (using either small, mid, large-sized retailers or its own retail network)
- Changes in marketing strategies (to focus more on a specific region or product)
- Mergers and acquisitions of new businesses
- Integration & closures of branches in a business
- Changes in competitors’ network strategy .

What we ultimately expect from a network optimization is cost reduction and service improvement. Although the level of importance among these two may differ depending on the purpose of the optimization, a network optimization, in the end, is to find the best combination of warehouses to offer a desirable service with the lowest cost possible under given constraints.

We may have to compromise a bit due to the inherent trade-offs between cost and service. However, we can find the best possible combination by analyzing and evaluating many different scenarios.

2. Types of Logistics Network & Project Methodology

Types of Logistics Networks

Point-to-Point vs. Hub-and-Spoke

In recent years, the industry has transited from a point-to-point model (P2P) to a hub-and-spoke model (H&S). In the H&S model, each hub in different areas is in charge of providing services to customers in the area. In the P2P model, however, a single facility serves the needs of customers across different areas. And it is been proved from experiences that the H&S ensures a better and more efficient use of one’s limited logistics resources.

In the P2P network, because a single manufacturing facility serves orders from each individual customer, it often causes long lead times and difficulties to reflect different product mixes from customers. Especially, if demand for a product is low, it becomes much harder to make production and shipping plans. To respond to this, if you create sales hubs in the supply chain, as in the H&S network, a quick and efficient response becomes possible.

However, this is not to say that the H&S network is always better than the P2P network. In some cases, applying P2P can be more effective. For example, if you deliver goods directly from a manufacturing facility to customers, the P2P network is a better option. In fact, it is still highly recommended as it offers various advantages such as logistics cost reduction, shortening lead time and less costs for storages.

From a supplier’s perspective, you may want to reduce total logistics costs by taking a MTO (Make to order) approach to as many freights as possible, as long as conditions like large-scale FCL (Full Container Load)/FTL (Full Truck Load) orders or enough lead times are guaranteed. You would also like to maximize the amount of freights delivered directly by using large container ships and CPFR2) while reducing the bullwhip effect3). In this case, the best solution would be to use the H&S network as its main network strategy and combining it with the P2P network from time to time when necessary.

Horizontal vs. Vertical

The H&S network can also be divided into two types: One is a horizontal structure in which products are delivered directly from a RDC (Regional Distribution Center) in 1-tier. The other is a vertical structure in which products go though 2-tiers of a CDC (Central Distribution Center) and then a RDC. Based on the amount and frequency of orders, products are classified into three categories (A,B and C) to establish strategies in a way to minimize inventory storage.

To be more specific, this is to have a differentiated way of delivering products by classifying them into three classes - A, B and C - by the quantity and frequency of orders. A class and B class items are placed in regional distribution centers (RDCs) while C class items are placed in central distribution centers (CDCs). This two-way system is to maximize transportation cost reduction for A class and B class items and to create inventory risk pooling effect4) for C class items.

Meanwhile, small-sized and expensive items like smart phones are transported from a single warehouse to different parts across the nation to minimize inventory costs and resolve security issues through a quick mode of transportation (air).

However, the vertical structure can be effective only when one’s logistics control tower system is firmly in place since the network base can be either CDC or RDC, depending on the mix of products ordered or demand trend. On the contrary, the horizontal structure can be operated in a relatively stable manner because each RDC independently delivers all goods regardless of its ABC classification.

Methodology of Network Optimization Project

There are largely two methods in logistics network optimization, mathematical modeling and simulation modeling. Technically speaking, the “best optimal” scenario can only be produced through mathematical modeling. However, simulation modeling is mainly used to optimize logistics network because unlike the mathematical modeling, it enables us to consider various constraints we encounter in the real world.

Simulation modeling is generally conducted in three phases. In the first phase of analyzing the current status of logistics, you need to understand operation and marketing strategies, analyze processes and data, and calculate target volumes. In the second phase, you define modeling constraints, simulation standards, candidate networks and methods of evaluation. Then, you select an optimal alternative by evaluating what-if scenarios. During the final third phase, you make a plan on how to implement the optimal alternative from the perspectives of capabilities, process and system.

Note that there are things to consider during each phase. First, when analyzing the current status, you need to classify products by their hierarchy, then do grouping by product line, considering their characteristics. Delivery destinations also need to be grouped by zip code and address, and then by TR zone.

This is what is called aggregation in which you do grouping work to minimize errors in target volume calculation and to do simulations more rapidly by minimizing errors in simulations and reducing data loads on the optimization system. Also, through interviews with operation/marketing teams, you need to understand mid-to-long-term (3~5 years) sales strategies/objectives. This way, you will be able to recognize what market share or demand each region/product aims to achieve, which will then enable you to create network strategies more specifically.

During the second phase, simulation standards and constraints need to be defined. And it is important to prove objectively if the simulation method is appropriate through baseline modeling where the latest performance data has been applied.

Next, you do simulations with to-be scenarios by combining different numbers and locations of warehouses. Here, functions of the warehouses should be defined for each to-be scenario. By comparing various alternatives and evaluating qualitative and quantitative effects, an optimal plan is finally selected.

During the implementation planning phase, based on the selected optimization plan, you first calculate the size of warehouses needed by year and by network, review terms and conditions of the current warehouses and work on detailed relocation plans. You also figure out what kind of infrastructure and how much of an investment are needed to implement those plans based on to-be layouts and operation processes.

3. Cello NOS(Network Optimization System)

An Accurate and Efficient Optimization Simulation Tool

When simulating logistics networks, the complexity grows quickly if there are multiple sites and candidate locations, and it is extremely difficult to do simulation with Excel office only. Therefore, it is more desirable to use a simulation tool for optimization.

Samsung SDS’ Cello NOS suggests the most optimized solution for the current supply chain, using its own modeling and what-if simulation. By typing in pre-determined input data, it enables users to create an optimized network that best serves their needs.

Cello NOS also helps users make rational decisions by performing what-if analysis and simulations, quick and easy. It can even let them compare different scenarios conveniently as you can copy a scenario in a project to create a new one.

Users can upload and download all data on the NOS in MS Excel format, so they can easily create and analyze data depending on their purpose. Furthermore, it shows location information of where the networks are and how far they are with each other on Google Map to give users a better understanding of their supply chain network.

In conclusion, Cello NOS is a solution that suggests a new supply chain strategy by intuitively understanding the structure of one’s supply chain, coming up with realistically optimal alternatives and quickly analyzing their impact.

Example of How Cello NOS Syncs with Google Map

4. Logistics Network Optimization in North America

What to Consider

For more details, you can download a White Paper from Samsung SDS Insight Report.



▶   The contents are protected by copyrights laws and the copyrights are owned by the creator and Samsung SDS.
▶   Re-use or reproduction as well as commercial use of the contents without prior consent is strictly prohibited.