In an ongoing SCM project, we had to implement global ATP and supply chain planning for interchangeable parts. In the military supply chain, this issue seems quite common, so it’s maybe worth to write down our current findings and invite others to discuss. In this blog, I will explain our current SCM master data model.
At first, it was quite tricky to identify the relevant I&S data feeds. We deal with some distributed system landscape, where knowledge around parts interchangeability is decentral and widespread. We identified feeds from own engineering and maintenance systems, from the manufacturers, from kits and assemblies management and finally a variety of I&S information collected manually by the MRP controllers and procurement. The spectrum of materials spans over both consumables and spares. Material values span from low value commodities to complete aircraft replacement units. Not all, but most feeds are utilized at the moment.
Along with the SAP implementation, some of the material management was shifted to manufacturer part numbers. MPN tracking delivers another standard class of I&S information. The existing Supersessions and Form-Fit-Function equivalencies are just extended by MPN explosion. Thus, we found a quite standardized mix of one-way and two-way material master relationships to tackle with. These build up a relationship network composed of FFF nodes and SSC edges.
We were then quite successful with mapping these I&S relations into the existing SCM master data solution. Check out /INCMD/UI transaction in APO. In the pic below, you can see such a mix of FFF and SSC relationships modeled into a network. In SAP terminology, this is called a Product Substitution Procedure composed of FFF and SSC.
We use these master data in gATP and CTM planning. gATP traverses the network depending on the exact start node. CTM utilizes the network more generally as a set. When one-way relations are contained in the network, there are adjustments on how stocks are incorporated into the planning run. Basically, stocks of superseded material is counted less than 100%, since it’s only partially usable.
There was some minor development work around this solution using existing SCM Badi’s, but no modification of the APO standard code was needed so far.
In future, we plan to extend the solution towards gATP ‘by tail number’. Even in a distributed landscape, it seems feasible to map individual weapon system configurations into SCM and build an exact provisioning solution. Basically, the existing Badi’s allow you to switch the edges in the product graph shown. We can turn nodes and edges on and off by weapon system or even by individual equipment.
Remark, there are ‘many ways to Rome’. The ideas presented here are just one. I tried to focus on more general findings, which may also apply to other planning mechanics like SPP or A&D rotables planning. The master data model should be adaptable to other environments.