The shift from make-to-stock to make-to-order has significant implications for production planning and scheduling. A schedule-driven approach to supply chain management can produce both tactical and strategic benefits for manufacturers in a make-to-order environment.

By Dan Bridleman & Jeff Herrmann

The trend toward mass customization is gaining momentum. More and more markets are demanding products configured to the specific requirements of individual customer orders. For many manufacturers, the number of base products and options in their mix has grown enormously. Many production managers are finding that they must be prepared to produce millions of different product configurations.

At the same time, customers are expecting greater responsiveness in fulfilling their orders. Having been trained to expect fast warehouse delivery of goods made-to-stock, customers now want equally fast delivery of their goods made-to-order. The Japanese automakers' goal of a "three-day car" is no longer a pipe dream, but a realistic goal that can be achieved in our lifetime.

Not every market has completely switched to highly customized products manufactured or configured to order. But manufacturers are rapidly adjusting to this phenomenon, and the trend is clear. Advanced Manufacturing Research recently reported that 73 percent of the companies it surveyed now schedule their production based on actual orders rather than on a predefined plan.

The inevitable result is that manufacturers are changing the way they build products. Long production runs of identical goods are much less common than in the past. A production run in today's plant is more likely to include many distinct models, sizes, configurations, colors and option mixes. And each finished product is likely to have different labor content, machining or assembly specifications, materials requirements, shipping constraints and specified delivery date. This has a significant impact on planning, scheduling and management of the entire supply chain.


The impact of make-to-order
In a classic make-to-stock supply chain, the manufacturing cycle is complete before the customer's order is received. Customer orders are filled from finished goods inventory, which is filled by production orders generated through a planning process. The planning-driven, push orientation of the supply chain results in a relatively straightforward set of management processes and supporting systems. But it isn't very responsive to specific configuration requirements or sudden changes in market demand.

In a make-to-order supply chain, the manufacturing cycle takes place after receipt of the customer's order. This gives the manufacturer greater flexibility to satisfy the customer with a variety of products and specially configured options. However, the customer-driven, pull orientation of the supply chain conflicts with the manufacturer's need to plan capacity, materials and other resources in advance. This leads to more complex management processes and systems than in the make-to-stock case, requiring changes both in management processes and in supporting computer systems and tools. Following is a detailed look at several of the key differences in managing make-to-order supply chains.


An increase in the importance of scheduling over planning
Make-to-stock manufacturers are heavily dependent on having the right products in stock to meet customer demand. Because a customer order may be lost if it cannot be filled immediately from inventory, the manufacturer's primary focus is on anticipating demand (forecasting), planning to meet that demand (manufacturing planning), and managing the distribution network to put the right product in the right place at the right time (logistics, distribution resource planning). In this context, production scheduling is mainly a tactical exercise aimed at squeezing maximum efficiency out of the production facility. This tends to be true whether we are talking about stocking facial tissue on a supermarket shelf or the heavy equipment of an industrial dealership, as long as the predominant model is make-to-stock. If there is an overriding supply chain optimization concern, it lies in optimizing inventory placement to maximize the matching of shelf-stock to demand.

On the other hand, in make-to-order manufacturing, the focus is primarily on order execution. Because every finished good is built in response to a specifically configured customer order, the customer has full visibility into the entire cycle from order acceptance to delivery; and the manufacturer's main goal must be to execute that cycle as quickly and efficiently as possible. While forecasting demand and setting up efficient distribution logistics are still important, the primary leverage lies in making sure that materials, labor and machine capacity come together on the plant floor at the right moment to produce the customer's specific order in the proper configuration. In this context, production scheduling becomes the critical determinant of the factory's ability to service the customer and compete effectively.

Thus, make-to-order manufacturing, by its very nature, tends to be a scheduling-driven rather than a planning-driven activity. Planning systems tend to deal in aggregates -- aggregate demand, forecast product groups, aggregate capacity, nominal option profiles -- and, as a result, can't give detailed information about specific orders and individual configurations. Make-to-order requires visibility and control of the plant's ability to manufacture a specific order out of specific parts at a predictable time. This leads to a schedule-driven view of the supply chain with optimized, constraint-based production scheduling as the core of supply chain management. Forecasting, logistics and material management don't go away; but in a schedule-driven supply chain, these functions are driven by the scheduling process and the schedule becomes the primary focus for managing operations.


A need for good, executable schedules
If the production schedule is to be the focus for driving the entire supply chain and fulfilling the customer's expectations, then the schedule had better be a good one. Unfortunately, the traditional model for supply chain management systems has a well-known flaw. Most existing MRP systems (including MRP II) do not take enough production capacity, logistics, or other production constraints into account to accurately model which specific orders a plant can actually fulfill in a particular time frame. In make-to-stock businesses, this lack of control over the accuracy of the production schedule has often been considered acceptable because inventory levels and shipment quantities were sufficient to buffer the lack of precision. In a make-to-order business, however, it is a much more serious problem because (a) late shipments are completely visible to the customer, and (b) in a schedule-driven supply chain, perturbations to the tightly-packed schedule have a ripple effect which throws everything off.


A need for optimized schedules, not just feasible ones
Even where more comprehensive capacity models and capacity planning have been added to the traditional MRP architecture (i.e., MRP II), the computation is generally capable of finding feasible production schedules, not optimal ones. This is still insufficient in a make-to-order environment. The diversity of product configurations being manufactured in a make-to-order plant can lead to major efficiency problems if the high variability in resource requirements between products isn't balanced out. We have seen plants with as much as a 400 percent variability in labor and/or materials content between successive products in a work cell or on an assembly line. With this level of variability, optimizing the total production schedule is critical to achieving efficient operation of the plant and reliable coordination with upstream suppliers and downstream logistics.


Substantial differences in the constraints which need to be modeled
Because of the variability in the goods being produced, make-to-order puts a much greater demand on the constraints needed to develop a good production schedule. Most production schedulers are familiar with machine capacity constraints, setup constraints, sequencing constraints (e.g., don't send a Model Q and a Model W with Option 5 down the line back to back), etc. However, make-to-order plants benefit from considering a wider range of constraints, including:

• Supplier constraints and cost tradeoffs
• Workload balancing -- machine and labor constraints
• Marketing goals -- financial and market-driven priorities
• Customer priorities
• Customer promise dates
• Shipping and logistics constraints

Traditional supply chain management systems have not looked at such a wide range of constraints in scheduling or planning production. But in a make-to-order environment, all of these constraints and more need to be considered simultaneously in order to arrive at the best possible production schedules.


A reduction in the planning/scheduling horizon
Because the emphasis in make-to-order is on fast turnaround of customer orders, both the planning and scheduling horizons tend to be shorter than in make-to-stock operations. This puts even more pressure on supply chain management, since with a shorter planning horizon there is less room to maneuver in trying to create, balance, or adjust the plan or schedule.


Greater impact of production glitches and scheduling changes
Make-to-order supply chains are more sensitive to production anomalies and sudden schedule changes. There are few plants in the world where a schedule published at 7 a.m. is still perfect at 10 a.m. -- machines break down, parts are missing, orders are canceled, priority orders are inserted. However, in a make-to-order plant, the impact of these perturbations is more severe since every product on the schedule has a unique configuration which is probably not interchangeable with any other product on today's production schedule. Since the production schedule has been carefully constructed to balance and interlock all the relevant constraints, any change to the schedule has the potential to lead to substantial inefficiencies.


Greater importance of coordinating feeder lines and component plants
As a result of the emphasis on fast cycle time and tight optimization of the schedule against constraints, each customer order in a make-to-order plant generates a chain of events which leads from the main production line or plant back through any feeder lines which supply subassemblies. For example, the final assembly sequence in a tractor plant may be used to drive the assembly sequence on a dashboard feeder line or in a supplier's transmission plant. There is generally little buffer inventory, since each subassembly manufacturing operation is also driven by (and uniquely identified with) the customer order. This means that any change to the schedule affects not only the primary production facility, but the production schedules at the feeder as well.


More customer visibility and contact at the factory level
In make-to-stock, the factory is buffered from the customers by inventory. However, in a make-to-order plant, every finished good being produced is linked to a specific customer order, and every change in that order needs to be communicated to the customer. Conversely, changes originating with the customer require interaction with the plant (e.g., "Is it too late to change the configuration of this order?"). Therefore, front-line systems and plant personnel must be capable of dealing directly with customer-generated situations and problems. This is either a great opportunity for a manufacturer to enhance its strategic positioning in its market, or a disaster waiting to happen.


Results at Case Corporation
We have seen several reasons why make-to-order manufacturing puts significantly greater stress on the supply chain than a make-to-stock approach. However, successful implementation of a make-to-order strategy can yield tremendous competitive benefits for today's manufacturers. Using the right tools and systems, a well-managed make-to-order supply chain can result in a fast and dramatic payback.

Case Corporation is a $5.4 billion manufacturer and distributor of heavy equipment for agricultural and industrial use, headquartered in Racine, Wis. During the past two years, Case has embarked on a sweeping reengineering initiative aimed at improving its responsiveness to customers, reducing finished goods inventory in favor of make-to-order operations, and substantially improving the company's supply chain management systems to support the new policies. To facilitate these changes, Case embraced an integrated approach to supply chain management. Case reengineered its core business processes, made improvements to its existing ERP systems, and installed new computer-based systems for optimized order-driven production scheduling.

A key step in replacing the existing push manufacturing strategy with a pull strategy was to introduce the concept of a supply policy which guarantees the delivery of customer orders within a specified time frame. For example, when a customer in North America places an order for a Magnum Tractor, Case now acknowledges a delivery date within the stated service target. This commits the entire supply chain, including all of Case's suppliers, component plants and logistics partners, to execute the order within the defined cycle time.

To generate optimized production schedules that would meet the new service targets, Case chose OptiFlex from Optimax Systems Corp., Cambridge, Mass. OptiFlex is a planning and scheduling system that uses genetic algorithms to generate optimized production schedules from an incoming order mix in the face of complex supply chain constraints.

Case installed its first OptiFlex scheduling system in its Racine tractor plant in April 1996. Running on Microsoft Windows NT workstations, the system generates a build schedule that covers the entire make-to-order time horizon. It balances customer orders against the constraints identified in assembly, including labor content, model sequencing, equipment capacities, load optimization and marketing priorities. The OptiFlex system also considers logistics issues, such as when container ships leave various ports, and how long it takes to transport tractors from the plant to these ports.

Within two months after installation, the factory had improved its ability to meet promised delivery dates to a record 93 percent reliability. A few months later it was making its dates more than 99 percent of the time. Schedules are now so reliable that Case even invites customers to come to the factory to see their unit being built.

Case's first-pass yields (the percentage of units ready to ship as they leave the assembly line) also increased dramatically because optimized schedules helped eliminate part shortages. Instead of issuing material releases from the MRP system, the schedule now initiates supplier delivery schedules that are sent by electronic data interchange (EDI) to suppliers. The system can create a shipping schedule as detailed as the supplier wants -- calling out daily or even hourly requirements. Now that schedule reliability has increased dramatically and the schedule is stable, Case can work much more closely with its suppliers to fine-tune inventory and shipments, helping suppliers optimize internally and lower their costs.


The customer schedules the factory
The implementation of an order-driven supply chain management system can make a big difference in make-to-order manufacturing. It is even more powerful to schedule incrementally as the individual orders arrive on a transaction-by-transaction basis. In a make-to-order environment, this allows the factory to approach optimal utilization of its resources in fulfilling customer demand.

The concept of available-to-promise (ATP) is not new, but past ATP systems answered the question, "When can the factory fulfill this order?" by matching each incoming order against the remaining unallocated slots of a pre-defined production plan. Newer technologies are making possible capable-to-promise (CTP) order acceptance, a form of optimized available-to-promise in which each incoming order is checked against the entire set of constraints which determine the factory's capacity -- machines, material, labor, logistics -- to compute a promised delivery date which the factory can commit to meet. In a CTP system, a customer, sales rep, or dealer can enter an order, check it for configuration accuracy, query the factory for a desired or best-available delivery date, and then confirm a reserved position in the actual production schedule, often within seconds.

CTP is a dramatic improvement on traditional ATP because it shifts the supply chain management paradigm from a planning-driven to an order-driven model, and because it supports dynamic rather than static allocation of the available resources to meet customer demand. This moves supply chain management out of the tactical arena where the focus is on lower costs and greater efficiency, and into the strategic arena where manufacturers can gain market share and competitive advantage by responding more quickly to customer needs and offering a higher level of customer service.

The technologies for implementing CTP exist today -- fast optimization algorithms and constraint logic, transaction-driven client/server systems, flexible object-oriented architectures and data models. All the necessary pieces have been successfully field proven and are commercially available. Thus the next step in make-to-order supply chain management -- the vision of "the customer scheduling the factory" -- is ready to implement.

Copyright © 2020 by APICS — The Educational Society for Resource Management. All rights reserved.

Web Site © Copyright 2020 by Lionheart Publishing, Inc.
All rights reserved.


Lionheart Publishing, Inc.
2555 Cumberland Parkway, Suite 299, Atlanta, GA 30339 USA
Phone: +44 23 8110 3411 | br> E-mail:
Web: www.lionheartpub.com


Web Design by Premier Web Designs
E-mail: [email protected]