Have you ever tried to reduce the costs of your logistics network? Have you ever wondered if it’s optimal? Are the nodes of your network in the right places?
Often not associated with the core business of the company, logistics and distribution aspects are frequently put on the back burner and end up being a significant cost source.
Logistics is a substantial lever for improving a company’s competitiveness. Several options are possible, such as inventory management, warehouse automation, and, what interests us in this article, optimizing the logistics network.
Companies sometimes have dozens of warehouses, hundreds of customers, and thousands of products. What’s the best combination? How do you find the optimal network that minimizes logistics network costs? The human brain can’t answer such a complex question, so specific software needs to be employed. In this article, Conseil 2.0 presents its methodology to address these challenges. It is based on both industry expertise and the use of network modeling software, Supply Chain Guru.
Supply Chain Guru (SCG) is a tool that enables the modeling of the logistics network and provides optimizations to find the best model for your business. We will present the main features offered by SCG in two parts:
- Modeling the reference network, or baseline
- Network optimization
Modeling the baseline: a crucial step for any network optimization project
The more advanced a project is, the more costly modifications become. The same applies in our case. It is essential to model the baseline before proceeding to the next steps. Otherwise, the results of future optimizations will not be usable.
Baseline: Why and When to Build It?
Why
There are three main reasons why it is imperative to build a baseline before any network optimization project. The baseline aims to:
1.Verify that the model is consistent with the company’s data:
– Are the flows correctly represented?
– Are the modeled costs consistent with the financial data?
2. Serve as a starting point for future modeling.
3. Act as a reference point for comparing proposed scenarios:
– How do my transportation costs and warehouse costs evolve?
– How does my service level evolve?
When
A properly defined baseline ensures that we are moving in the right direction. It should be built:
– After the collection and validation of data
– Before constructing future scenarios
The collection and validation of input data is a fundamental step, though often overlooked. There is a common tendency to assume that our data is immediately available and usable. However, as we will see later, it is usually not enough to simply extract data from the ERP, gather a few Excel files, and load them into the software.
Depending on the complexity of the network, it generally takes between 3 to 5 weeks to collect the data, and another 3 to 5 weeks to build the model. Attempting to shorten this time may lead to higher costs during future optimizations.
Dats collection
The data required for building a baseline can be categorized into several groups, depending on the complexity of the optimization project. Generally, the data to be collected includes:
1. Products:
– Typology
– Weight
– Volume
– Price
– Other relevant attributes
2. Sites:
– Locations of customers
– Distribution centers
– Production sites
– Size and capacity of these sites
3. Financial Data:
– Operating costs of distribution centers
– Transportation costs
– Customer orders
– Other financial metrics
4. Inventories:
– ABC classification
– Inventory levels
– Safety stock levels
Depending on the project and the desired complexity, additional data may be collected, such as:
– Type of fleet
– Availability by distribution center
– Other pertinent details
Fundamental Steps in Data Collection
For each project, Conseil 2.0 adheres to four fundamental steps during data collection:
1. Identification:
– Determine what data is necessary for the project.
– Identify sources for each type of data.
2. Extraction:
– Gather data from ERP systems, Excel files, and other sources.
– Ensure data extraction methods are consistent and thorough.
3. Validation:
– Verify the accuracy and completeness of the collected data.
– Cross-reference data with financial reports and operational records.
4. Consolidation:
– Compile and organize data into a coherent structure.
– Ensure data is ready for modeling and analysis.
By following these steps, Conseil 2.0 ensures that the data collection process is robust and reliable, providing a solid foundation for effective network optimization.
1. Verify that all required data has been provided
The amount of data is often substantial, and it’s easy to overlook some. It’s recommended to regularly track the status of data collection.
2. Ensure the integrity of the data
Before modeling the baseline, it’s imperative to analyze the data to assess its usability. Due to various reasons such as different IT systems, specific local requirements, etc., data from different sites may have different nomenclatures and formats, lack coherence, contain erroneous information, and some data may also be missing.
3. Clean the data
Based on the analysis that has been conducted, one can then decide to add data, exclude some (carefully anticipating the impacts), or make assumptions to address shortcomings (which must be documented).
Here are some common issues encountered during data analysis and cleaning:
– Duplicate data
– Inconsistent nomenclature
– Variations in information granularity across different sites
– Multiple and inconsistent sources of information
– Mixing of inter-warehouse transactions and customer orders
4. Validate the data with stakeholders
The stakeholders who will validate the project should be involved from the outset to avoid surprises. Do they recognize their data? Are they in agreement with the assumptions that may have been made?
It can’t be emphasized enough. Data collection is as lengthy as it is crucial for the rest of the project. Neglecting it jeopardizes all future analyses.
Documenting the constraints
SCG operates on the principle of moving towards an optimal model. Constraints are used to reflect reality, such as business decisions, preferences to work with a specific supplier, fleet limitations, driver hours per day, etc. Constraints can, among other things:
– Be based on min/max values
– Involve volume or weight capacities, costs, time periods, or flows
SCG utilizes constraint tables so that the user can input them (flows, inventory, production, sites).
Building the baseline
Building the baseline begins by creating a new database where all collected data (from ERP, WMS, reports, etc.) is consolidated. These data, primarily historical orders, transportation, production or procurement records, and product characteristics, enable the filling of the six required elements to model the network:
– Products
– Sites
– Demand
– Policies for:
– Transportation
– Procurement
– Inventory
Note that inventory and procurement rules can be omitted if only modeling distribution center-to-customer transportation, for example. Similarly, if production is a key element of your network, it’s also possible to model production policies.
Model validation
It’s crucial to choose the right performance indicators, which should be done simultaneously with data collection. It’s necessary to verify the results of the baseline to ensure its compliance with reality. For example, if we know that warehouse XYZ delivered 97,000 tons of goods during the fiscal year, but the model indicates 98,000 tons.
Is this discrepancy acceptable?
What is our tolerance threshold?
How does this discrepancy compare to the tonnage of other warehouses, such as ABC and JKL? Is it negligible?
What effort is required to refine the model?
The model will never be 100% accurate. The question to answer is at what point can we consider the model satisfactory? This question depends on each project and must be addressed on a case-by-case basis.
During model validation, Conseil 2.0 collaborates with the client team to ensure satisfaction on both sides and achieve the first major milestone of the project.
The road to optimization and simulation
Now that we have built the baseline and the stakeholders have validated it, we can proceed to optimize the baseline and run scenarios, such as adding or removing warehouses, implementing cross-docking, consolidating flows, etc.
If the baseline is correctly established, these scenarios are relatively simple to model. Teams can then spend more time analyzing the results and thus maximize the time spent on value-added tasks.
Conseil 2.0 is an official partner of Coupa, the publisher of Supply Chain Guru. Feel free to contact us if you would like to learn more about our methodology or for any project related to optimizing your network. We would be delighted to assist you with your challenges.