The Supply Optimizer is a powerful solver with many key capabilities to model a variety of scenarios and product lines.
- Subnetwork ID is used to segment problems in a non-overlapping manner, so that the optimizer runs can be in parallel.
- Fair share allows for better use of common resources and ensures the right product mix across the network nodes.
- The optimizer balances costs related to production, transportation, storage, and so on.
- Aggregated constraints are advanced capabilities that can use product attributes to further control the key constraints.
- Because the optimizer can support both push and pull, it is well suited for seasonal products so the pre-production decisions are economical.
- Telescopic planning makes it possible that run performance is reasonable for planning in both daily and weekly buckets to support short-, medium- and long-term aspects in the planning horizon.
The aim of optimization is to minimize the total cost of the supply plan. The Supply optimizer achieves this by minimizing costs, such as:
Production, procurement, and transportation.
Non-delivery or late delivery of products to satisfy demands.
Falling below the target inventory level or going over the maximum inventory level.
Violation of manually adjusted values (for example, adjusted customer demand) or minimum values (for example, minimum production receipts).
Inventory holding.
Optimization is performed by transforming the supply planning problem into a mathematical model (a mixed integer linear program (MILP)). The output is a feasible, cost-optimized production, distribution, and procurement plan for the entire supply chain network taking into account constraints.
Optimization is performed by transforming the supply planning problem into a mathematical model mixed integer linear program (MILP). The output is a feasible and cost-optimized production, distribution, and procurement plan for the selected supply chain network - hence it does not matter to the engine if the optimizer is interfaced with TS or OBP for integration of inputs and outputs.
The optimizer enables cost-based planning. Independent of the use case, it always minimizes the total cost of the supply plan. During the optimizing process, it intelligently searches and considers all feasible solutions to find the most cost-effective one in terms of total costs.
A solution is feasible for the optimizer when it respects all the planning constraints, for example, the sourcing options and available resource capacities. A feasible solution can contain non-deliveries, that is, not fully satisfied demands or safety stock constraint violations, or can violate other soft constraints.
Soft constraints have violation costs assigned to them. The optimizer violates soft constraints if they conflict with hard constraints (for example, maximum lot-sizes), or if their violation is more cost-effective. For example, the procurement costs could be higher than the violation costs for not satisfying a demand.
Besides these scenario-defined costs, the optimizer generates internal costs for pseudo-hard constraints, for example, for the violation of the manually adjusted values. As these costs are very high, these constraints will not be violated if a feasible solution exists which respects them.
Heuristics and the optimizer can be used together in order to leverage each others' advantages.
The optimizer can create plans that are optimized according to the following criteria:
Profit maximization
Delivery maximization
Optimizer: Delivery vs. Profit
The following details are relevant for Profit Maximization:
The optimizer maximizes profit, where profit is the difference between total revenue and overall cost. Total revenue is calculated as the sum of the revenue of all products sold to market. The revenue of each product is calculated by multiplying its sales price (that is, non-delivery costs) by the quantity shipped to the customer.
If the total cost of a finished product (comprising production, inventory holding, and transportation costs, and the cost of all components) is higher than its sales price, the optimizer chooses not to deliver and therefore not to produce or transport this product.
As described in this lesson, the TS Optimizer is a powerful solver with many key capabilities to model a variety of scenarios and product lines and is certainly the best option to plan many of those scenarios.
What needs to be considered when choosing between the TS Optimizer and the Finite Heuristic are the very specific business requirements the solver needs to address.
Moreover, it should be evaluated if the end-to-end supply plan must be calculated in a single planning run, or if multiple planning steps are more appropriate.
For example, the rough-cut production planning can be created by Finite Heuristic in SAP IBP, then sequenced by Production Planning and Detailed Scheduling (PP/DS) in S/4HANA, and then back in SAP IBP the distribution plan can be created by Deployment run.