Customer order model

Pull-based ATP Models Pull-based ATP models perform dynamic resource allocation in direct response to actual customer orders. Models of this type can range from a simple database lookup to sophisticated optimization. The purpose of pull-based ATP models is to make best use of available resources (including raw materials, work-in-process, finished goods, and even production and distribution capacities) to commit customer order requests over a period of time across a supply chain. The time horizon in pull-based ATP models is usually so short that a company can assume full knowledge about the availability of production resources. Pull-based ATP models are responsible for matching complicated customer requests with diversified resource availability. The specific decisions usually involve which orders to accept and, for each order, what quantity and which due date to promise. 

Sub-model I The customer order model includes the customer placing order activities within both the DSC and ISC, which link the domestic customer in the DSC with manufacturer and the international customer with manufacturer in the ISC. In the sub-model, the processes A, B and C show the domestic customer order and the processes 1, 2 and 3 show the international customer order. In this sub-model, the domestic and international customer orders are produced. 

In this chapter, the DSC and ISC will be represented by mathematical modelling, with the uncertainties noted in Chapter 5. There are four sub-models including a customer order model, a production model, a raw material ordering and transportation model and a finished goods satisfying customer demand and transportation model. 

The mathematical formulation for the domestic customer order model and the international customer order model are presented in the following two sub-sections, respectively. 

Customer ordering simulating customer ordering model (Sub-model I). 

In the following, we explain the key issues in modeling business processes with a simple example from customer order processing. Eirst, let us outline the scenario ... 

Xiameter is not about growth. It is about improving operational efficiency and providing better service to a niche of customers. The company introduced a new business model in quadrant III of Figure 4.8. There are times when products are unavailable, but customers know that this is part of the business model. To make it work, customers order products in minimums (full pallets) and the organization has to stay disciplined with a focus on sticking to the business rules. 

Cheung et al. (2008) developed a mathematical model for dynamic fleet management that takes into consideration dynamic data, such as vehicle locations, travel time, and incoming customer orders. The model is able to efficient re-optimize the route plan as dynamic information arrives, and it includes a genetic algorithm procedure for solving the static vehicle routing problem, and a quick heuristic procedure for dynamic updates of the vehicle routes as new data arrive. 

A company can also use different models in its own SOURCE, MAKE, and DELIVER links. For example, it can follow an MTO model in its MAKE link while its raw materials link follows a SOURCE STOCKED PRODUCT model. The make-to-order sandwich shop does just this, replenishing components to forecasts of daily volume (SI SOURCE STOCKED PRODUCT) while manufacturing each sandwich in response to customer orders (M2 MAKE MTO PRODUCT). 

Linked supply chain companies can also have varied models within their networks. For example, an MTS company might sell components to a MTO company. This would likely be the case with Dell component suppliers. The suppliers use the MTS model to quickly provide components on demand, while Dell uses the MTO model to fulfill customer orders. 

The Dell Computer Corporation is famous of its direct business model (see Dell and Fredman, 1999) that assembles customized computer systems based on customer orders, and ships directly from the factory to the customers. Based on Kraemer, et. al. (2000), Dell basically segments its customers into Transaction, Relationship, and Public/Intemational customer segments, in which more refined sales channels, including Home Home Office, Small Business, Medium Large Business, State Local Government, Federal Government, Education, Healthcare, are offered at Dell. By doing so, Dell is able to satisfy special needs in each customer channel. 

Push-based ATP Models In anticipation of future potential customer orders, a push-based ATP model pre-allocates lower-level availabilities (usually resources, including material, production, distribution and so on) into higher-level availabilities (e.g., finished products at a certain location in a certain time period) based on forecasted demand. The resulting allocated availability (or... 

Push-based ATP models span a relatively short time horizon, usually several days or a few weeks into the future. Within this time period decisions are based principally on a mix of confirmed customer orders and short term forecasts. In contrast, traditional inventory control and planning models cover longer time periods and decisions are based completely on (longer-term) demand forecasts. 

The alternative to rule-based models and allocation policy is the use of optimization-based resource allocation models. Such models can explicitly take into account variations in profitability at the customer order level, com-... 

Pull-based ATP models are executed in response to one or more customer orders. In an MTS production environment addressed in Section 5.1, the problem to be solved is one of assigning customer orders to some product availability pool, which could be a location of existing inventory or a planned production batch. The simplest optimization models of this category can be viewed as an assignment/transportation problem. In an ATO/MTO/CTO environment addressed in Sections 5.2 and 5.3, the problem to be solved is more complex. It involves assigning customer orders to appropriate manufacturing and material resources. 

In a complex supply chain network, multiple business units including retailers, distribution centers and factories may be considered in an ATP search. These business units can be dispersed geographically. Hence, an intuitive method for ATP search is based on distance (or lead-time) between inventory location and customer order location. The search process may also be based on a pre-assigned location priority scheme, or on the echelon-level of the business units, e.g. retailer level, distribution center level, manufacturer level. Considering a complex supply chain with multiple distribution centers and factories, Jeong et al. (2002) proposed a global search model to find avail-... 

More specifically, the models assume multiple products are offered. The materials needed in assembling end-items are grouped into different component types. The manufacturer may have multiple suppliers to provide materials of the same type, which differ in features such as quality, price and technology. The authors denote the combination of a component type and a supplier as a component instance, which represents the very basic material element in the models. Each customer order has an associated BOM that specifies the quantity of each component type required to build the customized product. Furthermore, for each selected component type, a customer can specify a set of preferred suppliers. The manufacturer is allowed to take advantage of this customer preference flexibility relative to components. This implies that the models only allow component substitution at component-instance level. However, the manufacturer has to take into consideration the incompatibility between certain pairs of component instances. This further complicates the formulation of these advanced pull-based ATP models. 

The two other kinds of flexibility associated with customer orders in the models are order quantity range and delivery time window. As mentioned earlier, the general order specification regime allows an order to include an allowable range on the desired order quantity and an acceptable range of delivery due dates. If the manufacturer cannot promise a quantity and a delivery time within the respective windows, the order is denied and assumed lost. 

As an order promising and order fulfillment engine, a pull-based ATP model is responsible for quoting a committed quantity and a due date for each order, for scheduling production to fulfill promised orders, and for configuring finished products at the component instance level. As mentioned earlier, customer orders are collected over a batching interval, the time lapse between successive ATP executions. The major decision variables in the advanced pull-based ATP models include ... 

They include the user as an integral part of the supply chain control system in their research test-bed. To this end, the simulation component dynamically generates live data, such as customer orders, to the ERP and SCM components. Then, the user makes decisions about the ERP and SCM systems to manage the supply chain. The system periodically gives feedback to the user and the rest of the model. Ball et al. intend to investigate both the performance of supply chain software and human decision-making. For more information about the model and how it relates to available-to-promise (ATP) decision models, refer to chapter 11. 

---