Structure of Advanced Planning Software Systems
In the last years several software vendors have developped and started to sell so-called Advanced Planning Systems (APS). They claim to support the integrated, capacity-focused and optimal planning of operations in complex global supply chains.
Considering early promotional publications distributed by software vendors, it appears that the structure of these so-called Advanced Planning Systems (APS) has developped historically with a two-dimensional view of planning as depicted in the following figure, one dimension being the business function involved and the other dimension being the planning horizon.
This planning structure, which seems to be the basis for most of the APS, has several
|1.||The multi-location-based structure of the planning system, which is required for Supply Chain Management, is not visible|
|2.||The problems treated by the different planning modules are not clearly differentiated with respect to the type of production and logistics system considered. However, it is known that the planning problems faced by, say, a metal parts manufacturer producing with small lot sizes in a job-shop environment are different from the planning problems of a manufacturer who uses a fixedposition layout and project-oriented production processes. This difference must also be reflected by the planning models used to support decision making.|
|3.||Material requirements calculations (MRP) are assigned to the short-term planning phase of the purchasing function. Obviously, MRP is not confined to purchasing. Even classical ERP systems associate material requirements planning with the production function. Developpers of APS seemingly are not aware of the problems inherent in the classical MRP approach. Therefore, APS do not replace the MRP planning phase with a better solution approach, but they still rely on the classical MRP calculations. Allthough it is well-known, that the generation of feasible production plans requires an integrated view of lot sizing and material requirements planning, this problem is still completely neglected|
|4.|| Roughly speaking, the planning problems emerging in a supply chain environment can be assigned to three layers: network design, node design, and operational planning.
Network design is part of a company's strategic planning activities, which are concerned with defining the long-term objectives and the principal course of actions required to meet these objectives. This includes the building of strategic alliances with key suppliers (and key customers) as well as the definition of the locations of all nodes making up a supply (and delivery) network.
Node design problems refer to the design of the infrastructure that must exist within a node in order to ensure the node's functionality within the supply network. For a production node (factory) the overall layout as well as the size (capacities) of the different production departments (e. g. flow lines, FMS, job shops) must be defined. Most of these node design problems are currently not supported by the APS available on the market. Therefore the corresponding layer is not depicted in the above figure.
Operational planning problems are concerned with the efficient performance of the detailed processes in a supply network. In this planning layer, which is the main application domain of APS, decisions concerning order fulfillment and resource utilizations are made. In other words, operative production planning is confined to the economical usage of the resources under the constraint, that their capacity is more or less given.
|5.||The module “Demand planning” is assigned to the same (hierarchical) planning level as the other planning functions. Obviously, forecasting (which is the same as demand planning) is not planning as it is not based on a decision model. The function of demand planning is the generation of input data for the other planning functions. Therefore it does not fit into a matrix with the two dimensions shown.|
|6.||Like forecasting, the module “Available to promise” as currently implemented by the software vendors has nothing to do with planning and does not fit well into the planning matrix. Up to now, it is only a data collecting function supporting several other functional areas in a company, e. g. the marketing department. From a more fundamental point of view, the complete column entitled “Demand” should be placed outside the body of the figure in a supporting module.|
|7.||Transportation planning is assigned to the short-term planning part of the distribution function. However, especially in global supply networks many operational transportation tasks have to be planned and performed with tight integration with the supply and production function. It is a specific characteristic of supply chain optimization that purchasing and production decisions are made simultaneously with the corresponding transportation decisions.|
In addition, planning is based on a deterministic view of the problems to be solved. Stochastic aspects are not or only insufficiently taken into account. As to our knowledge, currently the only stochastic aspect considered is the calculation of safety stocks. This calculation is taking a rather simple perspective of uncertainty which is basically the same as in the standard MRP systems. In addition, it is often assigned exclusively to the distribution function or the demand planning function. Uncertainties caused by breakdowns of resources and due to the variability of processing times and flow times are not handled at all.
|Tempelmeier, H., Supply Chain Planning with Advanced Planning Systems, in: Proceedings of the 3rd Aegean International Conference on Design and Analysis of Manufacturing Systems, 19.-21. Mai 2001, Tinos Island, Greece PDF-Download|