Hi Mohamed,
When using supply planning, you can specify additional parameters such as Minimal Lot Size and Incremental Lot Size.
When you model transport, for example, Transport Quantity is based on lot size parameters. There is a Minimum Lot Size (TMINLOTSIZE), and an Incremental Lot Size (TROUNDING). Incremental Lot Size is sometimes referred to as "rounding". These are enabled for horizons specified in the Optimizer Profile (TINCLOTSIZE is used by Inventory Optimization only).
Minimum and incremental rounding values are not time-dependent master data attributes.
For more information, I can recommend the SAP IBP Supply Optimizer Scenario Guide - Community Wiki:
"To plan supply chain scenarios the engine is creating and solving a linear program. With this kind of representation all variables are continuous. They can have any real value and any fraction. For many use-cases that’s perfect. But often you want to express that you handle pieces or that you want to transport, produce or purchase products in batches of a certain size, you define lot sizes. To solve such scenarios with lot-sizes the optimizer is still starting with the linear program and computing a solution with continuous real numbers. In the following steps the optimizer is “rounding” the solutions to integer numbers representing the lot sizes. But this is not a single easy step. Instead you need to check if the nearest rounding is the best solution or not. Assume the example we have a fractional solution of 4,3 and we must round it to pieces (rounding value 1). Then the optimizer might come up with three options: 4 as the nearest solution; all solution >= 5 and all solutions <=3. To see, what is the best value, the optimizer has for each three alternatives to evaluate, what is the best possible solution under the additional constraint =4, >=5 or <=3. To do so the optimizer must consider all the other lot sizes and must check all combination of cases, until the best solution is found. This leads to large search tees, which usually need much more time to solve than the first linear solution with fractional numbers. Modern solvers like Gurobi in IBP have many techniques to speed up this search process as good as possible, nevertheless the number of integer variables for lot sizes dominates usually the runtime. In IBP you can define on the planning profiles for how many periods you want to consider the minimal and the rounding lot sizes to control the runtime of the solving process. To show the impact of the lot sizes we show in the following scenario descriptions, which time is needed to find the linear fractional solution and how long the customer was running the optimizer to find the best-rounded solution. If lot sizes are important for a large horizon, usually the product-decomposition is used to split the overall scenario in small pieces and to do a perfect rounding for these small sup-problems. Here is the idea to focus on this decision locally, without computing every little effect of the lot sizes on other parts of the supply chain, even if it might exist. Beside the pure number of variables, the following scenario descriptions will also contain the number of binary and integer variables needed to control the lot sizes. Try to keep the horizon in which lot sizes are considered as short as possible. Especially if you do midterm calculations, rough simulations or S&OP-Planning try out the performance without lot sizes and think about the importance of the lot sizes on your planning granularity."
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.