Executive Briefings

Building an Integrated Model for Supply Chain Cost Optimization

Supply chain managers constantly struggle to optimize their supply chain costs. The key challenge lies in knowing where to create change based on relevant supply chain data. This paper outlines an approach to enable businesses to understand which optimization tool is the best fit to reduce costs and increase profits through supply chain management and provides a  strategy that combines two popular tools of supply chain planning - the Optimizer and Heuristics Planning.

Supply chain optimization is usually aimed at either reducing incurred costs or maximizing profit or both. This requires careful data analysis and demand forecasting, which can be self-defeating without relevant data. Manipulating variables to achieve effective supply chain management requires some tried and tested tools and strategies. In such scenarios, companies must choose the best fit solution to effectively optimize their supply chain.

SAP Advanced Planning and Optimizer (APO) Supply Network Planning (SNP) presents a model that enables companies to achieve supply chain cost optimization by effectively managing inventory. This module offers two different solutions that generate automatic replenishment plans for inventory - Heuristics Planning and Optimizer. The main function of both engines is to generate an automated inventory replenishment plan. This plan involves creating a receipt for all unfulfilled demands within the supply chain network.

At the outset, it would appear that the Optimizer solution is more popular than Heuristics due to its ability to handle a wide range of business scenarios. However, its complexity makes implementation and maintenance more time-consuming and less effective. Since it needs to access large amounts of data the Optimizer takes longer to create an optimal plan. If the execution is made time-bound the solution may not be properly optimized. Hence running the Optimizer at the lowest level of master data granularity can prove costly. On the other hand, Heuristics uses a simpler algorithm for its calculations making it a viable option at lower levels of data, although it may not satisfy all business requirements. It is important to note that Heuristics does not provide a cost-optimized plan apart from limitations in considering storage resources.

Challenges

Companies are often unsure of which solution - Optimizer or Heuristics Planning - to implement. In such a situation it becomes crucial to understand the specific advantages of each solution and how these benefits address the company's goals and overcome supply chain inefficiencies. To choose the right solution, companies must understand the challenges involved in the process of receipt generation employed and the type data analyzed, which are discussed below: 

Source Determination

Any product at a given location can be potentially sourced from various other locations. Source determination involves two key functions:

• Locating the correct source for receipt generation
• Selecting proper manufacturing methods for in-house products 

Heuristics enables source selection by defining any of the following variables:

• Priority -  each source location is given a static priority
• Quotas - each source location has a fixed quota
• Time - the source locations can be active during specific months of the year

Optimizing the supply chain to handle the situation illustrated above requires a solution that can handle the complex interplay of several variables. Heuristics is unable to execute such multi-variable high-level calculations since it works on static rules that can consider only one location at a time.

Tier Planning

Large supply chain networks have multiple tiers comprising clustered distribution warehouses and manufacturing units in different geographies. Businesses usually exhaust inventory within the cluster before placing inventory orders with supply chain factories. Heuristics cannot allocate inventory effectively where a cluster of warehouses function as a single entity in source determination. These capabilities are not out-of-the-box functionalities although external solutions can be created as supplementary tools. 

Automatic Inventory 'Holding' Location Planning

Most supply chain networks employ the hub-and-spoke model for distribution where a hub warehouse supplies inventory to several front-end warehouses. Effective supply chains store inventory at the hub warehouse and transfer it on confirmed demand to front-end warehouses, ensuring that the current inventory remains available to all units. Heuristics is unable to maintain such a process since it cannot perform intelligent grading for inventory holding locations. It transfers demand back on downstream warehouses instantaneously.

Choosing the right solution

Considering these challenges, it is useful to examine whether the company's supply chain faces frequent instances on chained source determination, tier planning or grading inventory holding locations. Usually these are exceptional scenarios that can be  addressed manually by altering the framework or by employing simple bolt-on solutions. The Optimizer requires a large quantity of master data maintenance as well as expert skills to decode its results. The optimizer tool is best applicable when the issues mentioned above crop up frequently.

Another approach is to classify the supply chain network. It is also possible that the above challenges occur frequently in certain parts of the supply chain network. In such cases, Heuristics can be chosen as the primary default planning engine due to its simplicity while the Optimizer is selectively used to tackle the localized issues.

A custom fit model that enables customers to create a minimum cost-sourcing solution using both Optimizer and Heuristics is the best solution. The Optimizer can run periodically to minimize supply chain costs and create time-phased quota arrangements. These in turn could subsequently feed regular Heuristic runs.

While this solution requires additional data and version maintenance within SAP APO, it can leverage the best of both Optimizer and Heuristics tools. Although setting up the cost data will be a requirement, the maintenance costs can be significantly reduced, thereby improving source determination. Such a solution can provide customers with a cost and capacity-bound time phased view of their sourcing/quota projections. Any change in costs can be identified by the Optimizer using real life transaction data that can provide realistic projection of the sourcing relationships. Since the usual supply chain planning is done using heuristics, it can avoid the challenges of frequent Optimizer runs.

Conclusion

In order to successfully reduce supply chain costs, companies require an effective strategy with relevant tools. The Optimizer and Heuristics tools have specific advantages in addressing supply chain inefficiencies and are commonly used by several businesses. However, each method has its disadvantages that can further increase costs. A custom fit model offers the best fit solution by integrating both Heuristics Planning and Optimizer tools within the supply chain network. By understanding and identifying the key challenges of chained source determination, tier planning and automatic inventory location planning, companies can understand how to leverage the solution to effectively reduce costs all the way across the supply chain.

Source: Infosys

Supply chain managers constantly struggle to optimize their supply chain costs. The key challenge lies in knowing where to create change based on relevant supply chain data. This paper outlines an approach to enable businesses to understand which optimization tool is the best fit to reduce costs and increase profits through supply chain management and provides a  strategy that combines two popular tools of supply chain planning - the Optimizer and Heuristics Planning.

Supply chain optimization is usually aimed at either reducing incurred costs or maximizing profit or both. This requires careful data analysis and demand forecasting, which can be self-defeating without relevant data. Manipulating variables to achieve effective supply chain management requires some tried and tested tools and strategies. In such scenarios, companies must choose the best fit solution to effectively optimize their supply chain.

SAP Advanced Planning and Optimizer (APO) Supply Network Planning (SNP) presents a model that enables companies to achieve supply chain cost optimization by effectively managing inventory. This module offers two different solutions that generate automatic replenishment plans for inventory - Heuristics Planning and Optimizer. The main function of both engines is to generate an automated inventory replenishment plan. This plan involves creating a receipt for all unfulfilled demands within the supply chain network.

At the outset, it would appear that the Optimizer solution is more popular than Heuristics due to its ability to handle a wide range of business scenarios. However, its complexity makes implementation and maintenance more time-consuming and less effective. Since it needs to access large amounts of data the Optimizer takes longer to create an optimal plan. If the execution is made time-bound the solution may not be properly optimized. Hence running the Optimizer at the lowest level of master data granularity can prove costly. On the other hand, Heuristics uses a simpler algorithm for its calculations making it a viable option at lower levels of data, although it may not satisfy all business requirements. It is important to note that Heuristics does not provide a cost-optimized plan apart from limitations in considering storage resources.

Challenges

Companies are often unsure of which solution - Optimizer or Heuristics Planning - to implement. In such a situation it becomes crucial to understand the specific advantages of each solution and how these benefits address the company's goals and overcome supply chain inefficiencies. To choose the right solution, companies must understand the challenges involved in the process of receipt generation employed and the type data analyzed, which are discussed below: 

Source Determination

Any product at a given location can be potentially sourced from various other locations. Source determination involves two key functions:

• Locating the correct source for receipt generation
• Selecting proper manufacturing methods for in-house products 

Heuristics enables source selection by defining any of the following variables:

• Priority -  each source location is given a static priority
• Quotas - each source location has a fixed quota
• Time - the source locations can be active during specific months of the year

Optimizing the supply chain to handle the situation illustrated above requires a solution that can handle the complex interplay of several variables. Heuristics is unable to execute such multi-variable high-level calculations since it works on static rules that can consider only one location at a time.

Tier Planning

Large supply chain networks have multiple tiers comprising clustered distribution warehouses and manufacturing units in different geographies. Businesses usually exhaust inventory within the cluster before placing inventory orders with supply chain factories. Heuristics cannot allocate inventory effectively where a cluster of warehouses function as a single entity in source determination. These capabilities are not out-of-the-box functionalities although external solutions can be created as supplementary tools. 

Automatic Inventory 'Holding' Location Planning

Most supply chain networks employ the hub-and-spoke model for distribution where a hub warehouse supplies inventory to several front-end warehouses. Effective supply chains store inventory at the hub warehouse and transfer it on confirmed demand to front-end warehouses, ensuring that the current inventory remains available to all units. Heuristics is unable to maintain such a process since it cannot perform intelligent grading for inventory holding locations. It transfers demand back on downstream warehouses instantaneously.

Choosing the right solution

Considering these challenges, it is useful to examine whether the company's supply chain faces frequent instances on chained source determination, tier planning or grading inventory holding locations. Usually these are exceptional scenarios that can be  addressed manually by altering the framework or by employing simple bolt-on solutions. The Optimizer requires a large quantity of master data maintenance as well as expert skills to decode its results. The optimizer tool is best applicable when the issues mentioned above crop up frequently.

Another approach is to classify the supply chain network. It is also possible that the above challenges occur frequently in certain parts of the supply chain network. In such cases, Heuristics can be chosen as the primary default planning engine due to its simplicity while the Optimizer is selectively used to tackle the localized issues.

A custom fit model that enables customers to create a minimum cost-sourcing solution using both Optimizer and Heuristics is the best solution. The Optimizer can run periodically to minimize supply chain costs and create time-phased quota arrangements. These in turn could subsequently feed regular Heuristic runs.

While this solution requires additional data and version maintenance within SAP APO, it can leverage the best of both Optimizer and Heuristics tools. Although setting up the cost data will be a requirement, the maintenance costs can be significantly reduced, thereby improving source determination. Such a solution can provide customers with a cost and capacity-bound time phased view of their sourcing/quota projections. Any change in costs can be identified by the Optimizer using real life transaction data that can provide realistic projection of the sourcing relationships. Since the usual supply chain planning is done using heuristics, it can avoid the challenges of frequent Optimizer runs.

Conclusion

In order to successfully reduce supply chain costs, companies require an effective strategy with relevant tools. The Optimizer and Heuristics tools have specific advantages in addressing supply chain inefficiencies and are commonly used by several businesses. However, each method has its disadvantages that can further increase costs. A custom fit model offers the best fit solution by integrating both Heuristics Planning and Optimizer tools within the supply chain network. By understanding and identifying the key challenges of chained source determination, tier planning and automatic inventory location planning, companies can understand how to leverage the solution to effectively reduce costs all the way across the supply chain.

Source: Infosys