Segmenting the Supply Chain
Until now, there have been few techniques available for analysis of something as complex as the typical business enterprise supply chain. We will discuss a new method for supply chain analytics based on dividing the supply chain into discrete functional units, called segments, which can be individually analyzed. After detailed customized analysis, the segments can then be linked to model any supply chain, from relatively simple to extremely complex.
Breaking the supply chain into segments provides a highly focused method of analyzing a product throughout the entire supply chain, from raw materials to the end customer. Segmentation allows companies to:
- Predict and proactively eliminate stockout situations;
- Evaluate actual and projected customer service levels;
- Dynamically model cycle times;
- Proactively identify and act on potential late shipments;
- Analyze supply chain requirements to meet customer expectations;
- Identify problem spots, bottlenecks, and congestion in the supply chain;
- Plan and implement appropriate safety stock levels and allocations; and
- Perform best- and worst-case performance analysis.
Customer Service and the Supply Chain
The success of a business enterprise depends on its ability to provide products and services to customers while remaining financially viable. Supply chain management seeks to provide quality products to the end customer at the appropriate time, in the correct amounts, and at the best price. Fierce competition and short product lifecycles have required the development of strategic partnerships among suppliers, manufacturers, distributors, retailers, and logistics companies. With these partnerships, it becomes imperative to shift the focus of supply chain management from individual company-oriented strategies to a methodology that optimizes the entire supply chain.
Meeting this goal requires integration of supply chain activities to focus on customer-oriented measures.
Logistics
Supply chain logistics involves the process of planning, implementing, and controlling the flow of materials from the point of origin to the point of consumption. Typically, materials undergo value-added steps (e.g., manufacturing) and/or integration with other materials at several points in the supply chain. In addition, temporary storage locations may be utilized to improve efficiencies or to provide a buffer against unexpected variations. Products are moved between these locations by transportation links. We define transportation links to include such logistic events as orders, packaging, shipment, receiving, and stock keep.
Segmentation
Evaluation of the supply chain shows that activities can be classified into two distinct functional areas:
- Inventory This operation includes holding inventory (warehouse, distributor, and retailer), performing value-added steps (manufacturing), or providing integration services (integrator, packager); and
- Transportation This operation includes ordering, packaging, receiving, stock keeping, and shipping of materials between locations.
Modeling With Segmentation
Modeling any path of a complex supply chain can be accomplished by combining appropriate segments. Each segment may be linked by noting that the ending inventory location of one segment will be the starting inventory location of the next. Segments can then be linked together to create a single path from supplier to end customer. Changes in the supply chain can be easily implemented by addition or removal of segments corresponding to the change. Similarly, custom segments such as rework (where beginning and ending inventory locations are the same) or internal operations (shop floor movements of inventory) can be created as necessary.
Flexibility in Analysis and Modeling
Because every segment structure is of the form inventory transportation link inventory, segment analysis can be made very modular. At the same time specific customized analysis routines can be applied as appropriate, without affecting other segments. Results of calculations are linked by defining the data requirements (input/ output values) and linkage path of the segments.
Variability in the Supply Chain
Segmentation of the supply chain not only provides flexibility in modeling and analysis but is also a key factor in reducing uncertainty. Standard inventory projection techniques, such as the economic order quantity, set safety stock levels based on variability typically measured and averaged at local points in the supply chain. These techniques result in relatively large uncertainties when projected to account for possible worst-case/best-case performance.
Segmentation of the supply chain provides focused analysis, utilizing incremental measurements within the segment. This allows projections to be made based on known times and locations, using detailed models specific to the segment. Therefore, projected uncertainty is associated with the partial path left to complete and specific knowledge of the segment events. Figure 2 shows the reduction in variability through segmentation and event monitoring. Variability projected from original order through stock keep (red line) is necessarily large to account for a wide variety of uncertainties.
Measurements of location at intermediate times (violet areas), combined with detailed characterization of steps, allow for significant reduction in overall uncertainty (tan areas).
Demand Variability
One of the most problematic uncertainties in analyzing the supply chain is the variability in customer demand. In addition to the standard issues such as seasonality, customer confidence, and economic considerations, traditional inventory management programs tend to create artificial demand variations sometimes referred to as the bullwhip effect. This variation arises when localized programs monitor direct orders to derive demand variability. The implementation of economic order quantity programs results in larger orders being placed infrequently to cover several days or weeks worth of demand. Of course, these large, infrequent orders are monitored by subsequent suppliers, resulting in even larger orders with higher variability progressing up the supply chain. This effect combined with other practices such as batch ordering, volume discounts, and stockpiling, results in high levels of observed demand variability and consequently large levels of safety stock in the supply chain. Eventually the supplier sees demand variability that has little to do with the actual customer demand but is highly modified by the separate inventory management systems within the supply chain. Figure 3 illustrates actual demand variability measured at the customer level and observed variability at the manufacturer level in a simple three-segment supply chain (supplier wholesaler customer). Customer orders are relatively stable and constant with some seasonal variation. Factory orders are more sporadic. Predicted factory orders using segmentation logic are shown by the yellow line. Total quantities are the same in all cases.
Using segmentation techniques creates a single path from the supplier to the customer. This means that actual customer demand can be used at any point in the supply chain. Aggregation of multiple demand requirements may be required for multiple customers using the same product. This type of aggregation is a simple statistical combination of variability (i.e., the square root of the sum of the squares of the individual customer variability). Batch orders and stockpiling may still occur, but the safety stock calculations will be based on actual customer demand.
Traceability Requirements
Detailed analysis techniques generally require traceability of individual components at all points in the supply chain. For most business enterprises, this is not feasible because of changes in product and responsibility throughout the supply chain. Segmentation allows easing of traceability requirements because different analysis techniques can be used at different locations. Thus, a segment with detailed traceability (e.g., serialized product movements) can utilize advanced analysis techniques while other segments with low levels of traceability (e.g., indistinguishable parts) can utilize averaging or queuing techniques, such as first in, first out (FIFO), to perform basic analysis.
Analysis
The segmented supply chain provides a natural customer-oriented focus by starting the analysis at the customer position and computing projected times and inventory levels for future customer deliveries. This type of analysis proceeds segment by segment toward the supplier, and is effective in predicting supply chain performance between the current time and the time that all replenishments are received (i.e., the event horizon is the lead time for all known orders).
The methodology not only identifies problems in the supply chain, but also estimates:
- Projected stockout situations;
- Projected customer service levels;
- Projected late shipments;
- Identification of congestion or bottlenecks in the supply chain;
- Time available before the problem is manifested;
- Potential impact of the problem on the customer; and
- Action necessary to alleviate the problem.
Segment Analysis
Because each segment can be analyzed independently, analysis begins with customer segment and current positions of demand, inventory, shipments, and times within the segment. Treating the segment as a virtual pipeline allows the software to compute:
Times for any shipment to complete; Expected demand for the time interval associated with the shipment (i.e., time from current position to completion); and Total estimated variability and required safety stock to provide required customer service levels during entire replenishment time.
Cycle Times
Monitoring cycle times for movements within a segment is critical to providing accurate predictions. Within each segment there will typically be specific events that provide information on product location and quantity. These events may include:
- Order Verification and allocation of product for customer;
- Shipment Pickup of product by carrier;
- In-Yard Arrival at destination by carrier;
- On-Dock Unloading of truck/rail car;
- Received Verification of product, entry into destination log; and
- Stock-Keep Product placed in inventory.
Monitoring these events provides measurements of actual cycle times, comparison of actual times to expected times with variability, and the evaluation of projected completion times.
Segment Demand
The inventory position at the end of the segment must maintain sufficient product levels to meet future demand while the next shipment is in transit. From the cycle time measurements made at the event level in each segment, a good estimate of the shipment arrival time can be made. This estimate can include both demand variability and cycle time variability.
Available Inventory
Inventory levels must be maintained at each segment endpoint to meet demand of the next segment or customer. This includes onhand inventory and future inventory currently in the segment (shipments or work in process [WIP]). The segment analytics must take into account the time required for any WIP or shipments to reach final inventory.
Safety Stock
Safety stock is a buffer stock that is used to maintain customer service goals when unusual events occur such as unexpected demand increases or shipping delays. The analysis of a segment computes safety stock requirements based on a targeted customer service level (i.e., percentage of shipments that can be met with on-hand inventory) and the projected variability in inventory levels. Projected inventory levels are based on current inventory combined with demand/variability and projected replenishment times/variability. Teradata SCI computes required safety stock for each position in the segment based on known shipment locations, customer demand, and variability.
Identification of Problems
Computing the inventory levels, replenishment times, and demand and safety stock requirements for individual event-to-event times within the segment, as well as the entire segment, allows easy identification of problem locations and suggested solutions. For example, a potential out-of-stock situation exists any time the projected on-hand inventory plus safety stock is less than the projected demand. Comparing each event in the segment provides the predicted location of the problem. From the known cycle times and demand models, the time before the problem arises and the shortfall quantity can be predicted.
Other calculations, such as bottlenecks, cycle time modeling, and actual customer service levels are performed in a similar manner.
Allocation Plans
Computing the required inventory levels as a function of time for the segment automatically verifies existing allocation plans, or can be used to generate future allocation plans by modeling.
Total Supply Chain Analysis
As analysis of each segment is completed, the required inventory, expected demand, and replenishment times can be passed to the next segment in the supply chain. Because demand and demand variability are directly attributable to end-customer demand, each segment will use consistent evaluations relative to customer needs.
Inadequate allocations or shipments identified in one segment will be passed to the next segment through the on-hand inventory and demand calculations. Thus, the root cause of any shortfalls or late deliveries can be automatically detected and traced to the origin well before it becomes an issue for the customer.
Conclusion
Localized inventory management and cycle time analysis increases inventory levels throughout the supply chain without significantly improving customer service.
Segment-based analysis is a new, effective way to tune inventory levels, lower costs, and increase customer satisfaction.
Teradatas Solutions for Supply Chain Intelligence utilize segmentation analysis and advanced logistics calculations to help companies:
- Predict and proactively eliminate stockout situations;
- Evaluate actual and projected customer service levels;
- Dynamically model cycle times;
- Proactively identify and act on potential late shipments;
- Analyze supply chain requirements to meet customer expectations;
- Identify problem spots, bottlenecks, and congestion in the supply chain;
- Plan and implement appropriate safety stock levels and allocations; and
- Perform best- and worst-case performance analysis.
