Order promising

In its pristine form, APS integrates three key processes advanced planning, advanced scheduling, and order promising. 

The third component of APS is order promising, which lies at the center of the transaction-based aspect of APS systems. This component is designed to suggest realistic promise dates for customer orders. The process, sometimes referred to as capable-to-promise (CTP), involves testing the customer s request date for feasibility and, if the date cannot be met, calculating the earliest date that it can be met. This is done based on available and projected inventory and available resource capacity. 

The transaction focus of APS systems renders their speed of execution critical to their effectiveness. Taking advantage of the many advances in hardware and software technology over the last decade, APS systems have execution speeds orders of magnitude faster than those of traditional MRP. Most APS engines have their data downioaded, either in batch or transaction mode, to a dedicated server that is architected to run memory-resident programs and databases. Under this scheme, they are able to deliver real-time order promising and make multiple runs to test various actions in an effort to further improve the plan. 

In the case of order promising, the demands on this timing change. Here, when completion date and an immediate response is expected. This is a synchronous process. Moreover, as new orders are entered, they need to be promised based on the latest information, including those orders that have just been accepted. Thus, the planning system, upon order acceptance, must immediately reserve materials and capacity so that aJl future promises can reflect the impact of even the most recently accepted orders. 

Keywords Available to Promise (ATP), Push-Pull Framework, Order Promising and Fulfill-... 

The fundamental business role of the ATP function is to provide a response to customer order requests based on resource availability. In order to make a reliable response to a customer order, an ATP system must insure that the quantity promised can be delivered on the date promised. Thus, an ATP system must include both order promising and order fulfillment capabilities. In addition, an ATP system should be able to dynamically adapt resource utilization and to prioritize the customer orders so as to coordinate supply and demand in a way that maximizes profit. By its very nature the ATP system should operate within a short-term operational environment where most resource availability is considered fixed because of procurement lead-time limitations. This distinguishes ATP systems from traditional plannings scheduling and inventory management processes. 

APICS defines ATP as ""The uncommitted portion of a company s inventory and planned production, maintained in the master schedule to support customer order promising Traditionally, the ATP scope includes the on-hand inventory and the planned production at a designated location. The MPS becomes moderately firm or even frozen once a designated time window is reached. This implies that the planned production quantity becomes static as the planned production time approaches. 

We now describe the ATP system for a particular electronic product (denoted by EP) manufactured by Toshiba Corporation. The EP supply chain consists of multiple final assembly and testing (FAT) factories all located in Japan, which provide EPs delivered directly to domestic business customers. An ATO production framework is employed. The order promising and fulfillment process involves several thousand product models. Order sizes range from a very small number of units to a few hundred. Orders are generated by one of several sales units and are processed by a single central order processing system. The ATP system collects orders over a 1/2 hour time interval and returns commitments to the sales offices at the end of each ATP run (V2 hour interval), order commitments are booked up to ten weeks in advance of delivery. 

The order-promising process employed partitions the due date time horizon into three intervals fixed product, flexible product and flexible resource. For the fixed product interval, which spans from approximately the present time to two weeks into the future, resources, in the form of manufacturing orders (MO) are fixed. An MO specifies the production quantity for each product at each assembly line in each factory. That is, a fixed production schedule is set, which takes into account both production capacity availability and critical material availability. Having a fixed schedule stabilizes production dynamics in the near term and allows for the required materials to be set up and put in place. Any order commitments made for this time interval must fit within the fixed production schedule. In the flexible product interval, two kinds of resources, capacity and material, are considered. The capacity consists of both production capacity in different factories and transportation capacity from factories to sales subsidiaries. The production capacity is given daily at factory level in terms of machine-hour and manpower availability, while the transportation capacity is specified as weekly maximum quantity from factories to sales subsidiaries. The weekly availability of individual critical materials is aggregated into finished good level availability grouped based on the bill of material (BoM). Any order commitments made for this time interval must satisfy the capacity and material availability constraints. The flexible product interval spans from approximately two weeks to two months into the future. For the flexible resource interval, which covers due dates more than two months into the future, the only constraint considered is production capacity. This interval starts beyond the resource lead times so any resource commitments can be met. 

In most transaction-based consumer channels, Dell uses a two-stage order promising practice that is widely adopted by e-commerce companies (see Bayles, 2001). The U.S. patented on-line configuration service (see Henson, 2000) at dell.com allows customers to configure their computer system by choosing options over CPUs, memory, operating systems, etc. As a result, Dell has to handle potentially thousands of possible configurations for each prod-... 

It is not unusual for the actual shipment date to be a few days different from the original promised shipment date due to uncertainty in supply chain processes. However, this unreliability might not be acceptable to relationship-based customers. Therefore, Dell offers a special service through Dell Premier accounts at premier.dell.com for business and public sector customers. The service not only provides reports and tools to assist purchasing, asset management, and product support, but also allows customers to hook up their ERP/EDI systems with Dell s to perform real-time computer systems procurement. Through Premier accounts, Dell commits to more responsive and reliable order promising and fulfillment solutions for its relationship-based customers with support from advanced IT systems (see Kraemer and Dedrick 2001). 

In one of its final assembly factories, Maxtor Corporation offers a component-level product. Hard Disk Drive (HDD), to its customers for further assembly of desktop PCs, high-end servers, consumer electronics products, network attached storage (NAS) server appliances, etc. (see www.maxtor.com). Maxtor has partnership-based B2B relationships with its customers. In this setting, customer orders are not given as specific order quantities and order due dates. Instead, only the total order quantity in each week is specified with permitted minimum and maximum quantity limits. In its order promising and fulfillment processes, Maxtor does not postpone customer orders to later weeks (i.e. no backorders), but it may deny customer orders subject to liability and/or penalties. While customer orders are promised weekly, the order fulfillment process is executed daily to provide accurate resource utilization and production schedules (see Ali, et al. 1998). 

Using ATP systems to accomplish order promising and fulfillment involves complicated modeling issues under different business environments. In this section, we systematically summarize the implementation dimensions, discuss factors affecting ATP implementations, and define push-pull ATP strategies. 

Customer response time - Customer response time measures the time lapse between order placement and the timing of the corresponding promise. Under an e-commerce relationship (especially associated with B2C), customers generally expect to receive an order promise (or response )... 

Order Promising Reliability - Customers under B2B and B2C relationships often have different tolerances for the order-promising reliability. Generally, in a B2B setting adherence to promised due dates and quantities is more important as production schedules and downstream processes can be adversely impacted by poor order fulfillment. 

ATP quantities) will be used to support future order promising upon actual order placements. 

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 ... 

The mixed integer programming models of Chen et al. (2000, 2001) includes constraints for order promising and fulfillment, finished product inventory, material/component inventory, material requirements, capacity utilization, and material compatibility. We highlight the major constraints as follows. 

ATP is an IT-enabled business practice that provides a response to customer order enquiries based on resource availability. It supports order promising and fulfilment, aiming to manage and match the demand to production plans (Kotzab, 2001). There are two types of ATP practices push-based and pull-based. The former is to compute ATP quantities and dates based on forecast demands, whereas the latter is based on actual customer demands. The ATP model supports... 

Venkatadri, U., Srinivasan, A., Montreuil, B. and Saraswat, A. 2006. Optimization-based decision support for order promising in supply chain networks. International Journal of Production Economics, 103, 117-130. 

---