Something you would not expect to see in a just-in-time work environment is

• Tweet

  Nội dung chính Show

  • Problems with Suppliers
  • Late Production
  • Need for Software
  • Control of Inventory
  • Low-Volume Operations
  • Batch Orientation
  • Management Complacency
  • Involving the Work Force
  • What are the problems of just
  • Which of these is not a goal of just
  • Which of the following are not the characteristics of the JIT philosophy?
  • What is a just

• Post

• Share

• Save

• Get PDF

• Buy Copies

• Print

Just-in-time production, or JIT, has probably received more attention in a short time than any other new manufacturing technique. The main reason is that JIT gets the credit for much of Japan’s manufacturing success.

Despite the extensive publicity and interest, few companies have implemented JIT in their manufacturing operations. If JIT provides all the benefits claimed for it, why have so few factories adopted it?

JIT’s widespread publicity has been a mixed blessing. The popular press, and even some technical articles, focus on the easily observable differences from batch production systems but ignore some of the more important but subtle features of JIT. Writers rarely get very far past the lower inventory costs attributable to JIT and seldom describe how the technique can improve the entire manufacturing process. Managers who have read only a little on JIT rarely understand how it can help their operations. Usually they focus on the fact that, in the end, JIT increases a company’s ROI.

More important than the reduction of inventory and greater ROI are the improvements in manufacturing that result from operating with low inventories. JIT removes the security blanket of high inventory and thus exposes related operating problems. These are problems that need to be faced and solved—and therein JIT can be seen to create hurdles of its own.

Converting to JIT means a big change—in the culture of a company as well as in its manufacturing operations. Established routines and rules become obsolete. Where backup inventories were once considered to be insurance against unexpected shortages or delays, they are now viewed as evidence of lack-luster planning or controls, even of laziness. Large production batches can no longer be viewed as beneficial because they help amortize setup costs. JIT forces the elimination of the waste inherent in long setups.

Few manufacturing organizations are very flexible, either in their operations or in the minds of their creators. A typical operation is like a huge steamship, for which a rapid change in course is difficult. Most factories have been making similar products using similar processes for many years; their managers are comfortable with what they know. In this environment, change comes slowly. This inflexibility combined with misperceptions of JIT keep a lot of executives from using JIT. They excuse themselves by saying: “I know JIT has done a lot for others, but our plant, and our processes, even our people, are different. In our situation, JIT won’t work.”

Since misperceptions create a roadblock to implementation of this valuable management technique, let’s look at them first.

Problems with Suppliers

Excuse number 1: “Our suppliers won’t support JIT by delivering our raw material in small batches on a daily basis.”

Asking suppliers to make daily deliveries is a common mistake of managers who focus on the inventory-reduction benefits of JIT. Ultimately, this is the right thing to do, but it’s the wrong place to start. If manufacturing executives recognize JIT as a problem-solving technique rather than an inventory-reduction plan, the proper starting point will be clearer. JIT should be adopted and practiced inside the factory, where the company can control any problems, rather than outside, where close cooperation with another organization is necessary. Once a company begins to master JIT, it should then begin to work with suppliers to help them understand the benefits it holds for both parties.

Furthermore, JIT is a demand-pull system. Each operation produces only what is necessary to satisfy the demand of the succeeding operation—in contrast with the traditional batch-push system, in which parts are made in large batches and pushed to the next operation on a fixed schedule. Ultimately, every activity in the factory and every demand on vendors is driven by the final assembly operation. Final assembly is the control point for the entire manufacturing process, and it is the place to start implementing JIT.

JIT demands that the production process be rationalized and simplified. For uninterrupted flows in a demand-pull environment, the schedule for final assembly must be smoothed out. As I said earlier, this schedule will drive every activity in the factory and thus will level the operating rate of each. Operating rates of different processes can then be matched, and buffer inventories that separate processes eliminated. All processes will operate better without the wide swings in demand characteristic of a traditional batch-push production system.

As companies implement JIT in this way, suppliers will find a predictable demand for their products as a result of the smooth final assembly schedule. Ultimately, they will also enjoy a consistent long-term demand. The prospect of being able to plan their operations will influence them to sign up as JIT suppliers.

In the long run, the factory will want only one supplier for each purchased part. This long-term supplier must always deliver a quality product, on time, in small batches. In return, the factory will give the supplier a long-term purchase forecast with a guarantee to buy at a percentage of it. In addition, when the JIT operation is in place, suppliers will know that the company is serious about the concept and can share the experience of implementing the technique as they become JIT suppliers.

JIT’s success depends on the high quality of incoming materials. If a supplier delivers a bad batch, the whole production line will stop! Once suppliers understand the consequences of failure, they will be sure to make on-time deliveries of high-quality materials. Although the relationship between manufacturer and supplier in a JIT setting entails risks, the rewards of perfect parts always delivered on time are tremendous.

At Hewlett-Packard’s computer systems division, just-in-time production was off to a good start before the procurement department made any attempt to convert suppliers to JIT delivery. Although most of our suppliers are still not making JIT deliveries, we are helping them to consistently deliver quality parts on time. Once they have met all the prerequisites for a JIT relationship, we will convert them to JIT suppliers. In the meantime, our efforts have resulted in improved quality, more frequent and smaller deliveries, and a reduction in raw-material inventory.

In our initial efforts to convert suppliers to JIT deliveries, we concentrated on the manufacturers of large components that require a lot of storage space. Although most suppliers of these parts were willing and able to cooperate, in one case we changed to a supplier who was closer and more responsive to our needs. We introduced the vendors to JIT concepts and then showed them their implementation on our own production line. This demonstration convinced them to supply our needs with JIT deliveries.

Making JIT deliveries did not mean that in every case our suppliers immediately adopted JIT in their own factories. One vendor of large metal frames continued to produce them in monthly batches in his own shop while making daily deliveries to us. After several months, when we told him that some of the frames were out of square, he realized his tooling had worn out without his knowledge. He also discovered a month’s worth of inventory of defective frames in his shop. Had he been producing frames just-in-time, he realized, he would have had to repair only one day’s worth. Besides improving his inspection procedures, the vendor quickly adopted JIT and asked us to help him sell the concept to his other customers.

After attacking the problem of excessive large-part inventory, the computer systems division began working with suppliers of high-value parts. Most of them are distant vendors of unique parts. They were harder to convert to JIT. Most are still not operating on a strict just-in-time basis, but we have reduced our raw-material inventories, and multiple deliveries within a week are common. In addition, these vendors now ship more than half of their parts directly to stock without the need for incoming inspection.

In winning vendors over to our viewpoint, humor sometimes helps. Our materials manager explains that we understand how difficult it is to produce perfect parts and that we will accept a few rejects from our suppliers. We insist only that our suppliers first separate their bad parts from their good ones and ship them to us separately. Furthermore, if a delivery is going to be late, our materials manager requests only that the supplier’s vice president of marketing give our production workers a presentation on satisfying customers while they wait for the material to arrive. Suppliers get the point.

By working with our suppliers and showing them what we expect, we have reduced the average number of parts back ordered (past delivery date) from more than 200 to 2. Despite this success, we still have a long way to go in converting many of our suppliers.

Late Production

Excuse number 2: “We will always have back orders in our factory. We are constantly expediting production to make up for these shortages and to complete products for shipment within the scheduled cycle time. If we go to JIT, the line will always be shut down, and our production will always be late.”

This is a common lament of production managers who feel that they are always making up for the poor performance of the materials department. In their view, although late deliveries of incoming materials or subassemblies are undesirable, late shipments of finished products are unacceptable. On the other hand, while the materials department tries to make deliveries on time, it knows that products can be made in much less than their scheduled time and it often helps the production manager expedite late orders so that shipments will not be missed.

With long production cycles and large in-process inventories, this kind of adaptation will continue indefinitely. With large in-process inventories, the actual time spent working on a product represents a small percentage (often less than 5%) of the production cycle’s length. During most of the cycle, a product sits in inventory waiting to be worked on. Although expediting enables products to bypass these inventory queues, a large factory requires an army of expediters.

With JIT, large inventory queues don’t exist, so the production cycle can often be reduced by 90% or more. Then production schedules really mean something, and expediting is no longer possible. When production schedules represent the minimum time possible for a product to move through manufacturing, the materials department will understand that late deliveries mean late product shipments. Unwilling to be responsible for late product shipments, the materials department will strive to eliminate late deliveries. And when buyers or schedulers understand that a late delivery will stop production and generate a crowd of anxious managers around their desks, they will devote more effort to ensuring that materials arrive on time.

A safe method for introducing JIT into this type of environment is to leave the cycle times of the materials planning system intact while shortening the actual production cycle. The effect will be to move the inventory from the factory floor back to the stockrooms. While this strategy will not reduce inventory, it will improve the flow of materials through production. Indeed, that has been the objective of traditional queue-reduction programs. Parts that are late from suppliers will usually be delivered before production requests them from the storeroom.

With inventory pushed back into the storeroom, late deliveries will cause fewer problems and less confusion on the factory floor. Parts delivered after they are needed by production will cause late shipments. The results, however, will be no worse than they would have been under the old system. If there are no late product shipments under a batch system, there will be none under JIT. As deliveries improve, cycle times for materials planning can be matched to the production cycle, and inventory will be reduced.

Need for Software

Excuse number 3: “Our batch-oriented materials planning and control system won’t allow us to operate in a just-in-time mode. We need to install a just-in-time software package before we can convert our production operation.”

Experienced managers know that forcing a production process to fit a software system is a prescription for disaster. Designing processes to conform to the requirements of a particular software package often makes operations less effective. The process needs to be converted first.

Before installing a software system, a company naturally should understand the process that the system will be supporting, how the system should be designed, what features it should have, and how it should operate. This can be done only if the process is designed first and has been debugged through operating experience. This is especially true when implementing JIT because it is so different from a traditional batch-push operation. If manufacturers’ basic philosophy and operations do not change, workers will figure out informal ways to get around any JIT computer system that may be installed; for example, they can accumulate materials at their individual workstations to serve as buffers.

What then can be done in the period between the start of JIT production and the implementation of a JIT computer system? How can managers be sure the batch materials system accurately reflects the status of materials in production?

There are many ways to synchronize a batch-oriented system to JIT production. The techniques depend on the most flexible system resource ever conceived—manual clerical effort. Obviously, better long-term solutions exist than reliance on pencil and paper, but when both information needs and production processes are changing rapidly, nothing is more flexible.

Simple forms can be designed and used to monitor the flow of materials on a piece-by-piece basis. Inputs can be made to the system in batches when a form is completed. Materials allocations automatically made in batches by a traditional MRP system can be manually adjusted for individual pieces. Japanese-style kanbans (or American chits) can be used to requisition parts on an individual basis, and systems transactions made only when a certain number of kanbans are accumulated.

These are just a few examples of how a manual system can temporarily serve in place of a formal materials planning and control system. The technique will depend, of course, on the operation it supports and will probably change as the process evolves. Once the operation has stabilized, the company will have a good understanding of the requirements for a materials system and will know how to customize a software system to fit its operation.

A hard-line adherent of JIT might say that a company doesn’t need a computer system for materials planning and control because materials are pulled through the factory by demand. This demand is transmitted back to vendors, and materials are pulled into the factory’s first operation. Although some Japanese companies do operate within guidelines that approach this theoretical limit, U.S. manufacturers will probably never feel comfortable with this type of operation. Although they may be able to replace complex computer systems with simpler ones, few will care to give up the planning capability they have today for a system driven by the often erratic demands of the marketplace.

Control of Inventory

Excuse number 4: “If we adopt just-in-time production, we won’t be able to track materials through the factory with work orders. So we’ll lose control of our inventory.”

Accountants, production control schedulers, or expediters usually raise this objection. They are the ones who operate (and may have developed) an elaborate inventory control system. With hundreds or even thousands of batches of material in process, a system for tracking work orders is necessary to prevent chaos. Eliminating such a system in a traditional environment would mean a total loss of control.

If production is simplified, however, the tools required to monitor and control it can also be simplified. In a JIT setting, little inventory is on the floor, the flow of materials is clear, and the production cycle is short. Thus even without tracing work orders, accountants and schedulers actually have more control than in a traditional system. Inventory consists primarily of either raw materials or finished products, both of which are easy to count and value.

At Hewlett-Packard’s computer systems division, the finance department was worried that control over inventory would be lost under JIT, and schedulers responsible for planning subassembly production feared they would lose the production tracking capability they felt was necessary for their work. To satisfy the finance department, we conducted an inventory of work in process at the beginning of each month. Because work in process was expected to be low, the inventory could be taken quickly. In the first month of JIT operation, the inventory took two hours. We have since reduced it to a half-hour exercise.

The subassembly production schedulers gave up their traditional scheduling method as no longer necessary. Now they do not need to maintain particular inventory levels within in-process line stocks as buffers between production operations. Calculations of safety stock and safety days no longer determine the right inventory level.

Low-Volume Operations

Excuse number 5: “We are a low-volume operation, so we couldn’t benefit from JIT.”

Most production operations involve low volumes. The principles of JIT are just as applicable here as in high-volume operations. Emphasis on reducing setup times, building products in smaller batches, and making things only on demand will improve a small operation as much as a large one.

In fact, a small-volume operation may find it easier to convert to JIT because it may already be making products in small batches using simple equipment with short setup times. Hewlett-Packard’s computer systems division, which builds fewer than six computers a day, has realized tremendous benefits from JIT. These are, however, complex machines with many subassemblies and thousands of individual parts.

Batch Orientation

Excuse number 6: “We’re a job shop, so our business is naturally oriented to batch production. We can’t use JIT.”

If a job shop got only unique orders whose patterns were unpredictable, just-in-time production would meet with little success (although reduced setup times could still be beneficial). Most job shops, however, have much more repetitive business than they realize. By not taking advantage of this fact, they are losing the chance to improve their operations.

As a first step to such improvements, job shops should separate their repetitive business from their unique orders and develop special production methods for the repetitive work. Equipment can often be dedicated to particular tasks as a way to eliminate the unproductive setups required to change machinery from one batch job to the next. A JIT production line set up to do this work will be much more efficient than all-purpose equipment in a batch environment. If demand for a particular product is insufficient to dedicate equipment to its fabrication, groups of dissimilar equipment in manufacturing cells can often be dedicated to making families of parts that require the same manufacturing sequence.

Use of “group technology” allows parts to be made in small batches (or even one at a time) with short setup times. The cycle time for producing a part in a small cell of closely spaced machines will be much shorter than it would be for a large batch that has to travel around the factory. The quality feedback for this cell will be much quicker, and the production operations can be more finely tuned.

Management Complacency

Excuse number 7: “Our factory is operating okay already. We don’t need to put in the effort to convert our operations to JIT.”

This is the most dangerous excuse of all but probably the most prevalent. It demonstrates the complacency that has already been the downfall of manufacturing in many U.S. industries and that threatens many more. Our international competitors are improving rapidly, and if we want to stay in business, we must keep pace with them. Waste and confusion in our factories harm both productivity and product quality. Converting to JIT manufacturing is an excellent way to expose problems and improve operations, but it requires top management commitment and effort.

Although commitment is always mentioned as a prerequisite for any organizational change or new program, it is especially needed when converting to JIT for two reasons. First, JIT exposes manufacturing problems. To develop an excellent factory, managers with the authority must be willing to commit resources to solving these problems. The best candidate for conversion to JIT is the organization that has a quality-control program and is already documenting its processes, measuring their performance, and eliminating problems. The computer systems division adopted statistical quality-control techniques two years before conversion to JIT began.

Second, JIT brings fundamental change to the organization’s way of doing things. When a production line stops because of a product defect, the emphasis in JIT is on determining the cause of the defect and then fixing the process accordingly. No longer do production workers attempt a quick fix; instead, the focus is on continuous improvement. Workers must think not just about doing their jobs, but about doing them better. Thus training is needed to reduce fear and to enlist the cooperation of the entire organization.

Involving the Work Force

Management commitment by itself will not ensure an efficient or rapid transformation to JIT. First, the staff must have the time to implement JIT. In any organization, everybody is usually occupied by their existing jobs, and no one has time to work on improving things. Making the extra effort will be low on everyone’s priority list, and progress toward JIT will be slow. The solution is to appoint a project leader who will champion JIT’s implementation.

In some organizations, the person who initiated or has been most favorable toward JIT becomes the champion responsible for setting up training, making a conversion schedule, and forming a conversion team with representatives from every department. The JIT champion will have to deal with the fact that no paradigm exists so that everyone can pursue the same objectives. The best way to overcome any problems is through a common learning experience.

Such training can be tailored to the different levels in the organization. For example, a group may attend professional seminars outside the company or internal classes taught by an expert in the field. Other useful activities include jointly reading and discussing a book on JIT, jointly visiting another factory that has implemented JIT, and creating models or simulations of JIT production lines that all can observe. At least one division of Hewlett-Packard has used all these techniques either alone or in combination. The Greeley Division has attracted some attention for the simulated production line it developed to show the differences between “push” and “pull” materials movement and the effect of moving materials in different-sized batches. The company has videotaped its demonstration of these effects, and the tape is available as the basis for a group learning experience.

Even after this conceptual training is complete and the organization has a common vision of the future, no company is prepared instantly to transform its entire factory into a JIT operation. To give everyone a better understanding of how JIT works in practice, the company should establish a pilot JIT project before converting the whole factory. In a plant with multiple products, this project can involve an entire product line (preferably not the largest). If it is impossible to separate out one product line or if the factory produces only one product, the pilot implementation can be done between two operations in the plant (preferably the last two).

The computer systems division, which builds only one major product, the 3000 Series 68 superminicomputer, adopted JIT between the last two processes in the factory. When employees felt comfortable with this arrangement, the division quickly implemented JIT back through the factory to the first production operation. Besides allowing an organization to experiment with various concepts and techniques before making commitments for the entire factory, the pilot allows mistakes to be made that are small enough to be easily fixed. Most important, it enables people to experience JIT in a real situation.

Visiting a factory using JIT is still another way to see the concept in action. If pictures are worth a thousand words, then observing another factory using JIT is worth several chapters in a book. Ideally, a company would visit a factory making the same product as it does, but if its competitors have already reaped the benefits of JIT, they will be unlikely to pass on their experience.

As the time for implementing JIT approaches, another problem will emerge: a small group of managers cannot anticipate all the many detailed changes that have to take place in production and overhead procedures. The obvious solution is to enlist more people in the implementation effort, and the right way to do this is to enlist everyone.

Nonsupervisory employees on the production line and in the office are the ones who know the most about how the system actually operates, and they probably have a lot of ideas on how to improve it. Their involvement will also make them more receptive to the change and will give everyone a greater sense of ownership of the new system.

Many organizations have recently begun to tap the tremendous potential in their first-line people by involving them in quality circles or quality-of-work-life teams. If such programs already exist in an organization, they can be built on in implementing JIT. If such involvement has not been a tradition, implementation of JIT presents a perfect opportunity to get started.

At the computer systems division, a team of managers outlined the new flow of materials through the factory. Then each supervisor and his or her people enumerated the details involved in building, testing, and moving one assembly at a time through each operation. Employees met often in teams with their supervisors and worked out the necessary operational changes.

When the changeover to JIT occurred, plans had already been made to accommodate the expected production problems. Actually, very few problems surfaced right away because the production workers had done an excellent job of planning. No group of managers, engineers, or planners could have done as well.

Staying flexible may mean having extra process engineering resources available to solve problems as they arise. It may also mean scheduling the production operation at less than its maximum capacity for a while so that changes can occur without compromising the factory’s scheduled output.

The last important step in implementing JIT should be to position the organization for continuous improvement. One method of doing this is to continue to remove inventory between operations in the factory. This reduction will not only expose operational problems but also allow production departments to move closer together, reduce the effort devoted to materials handling between them, and improve communication.

Eventually, all production employees could work next to each other in a linear (or U-shaped) production flow. One person would hand material to the person performing the next production activity. This would minimize inventory and materials handling, maximize communication, and result in better feedback on quality and ideas for improvement. In addition, the teams set up to plan operating procedures should be maintained and encouraged to focus on continuous process improvement. After all, people on the production line will have the best information on how the process is operating and what can be done to improve it.

JIT cannot be quickly put in place and forgotten. Implementing it is a commitment to operate in a new way, a better way—a way that demands things be done right. This method will not forgive inattention to solving problems. Above all, JIT is not just a way to reduce inventory in order to get a better return on assets; rather, it is a means of solving the problems that block the building of an excellent manufacturing organization. Some enlightened manufacturing managers have already made the commitment to JIT. If you haven’t started implementing JIT, what’s your excuse?

A version of this article appeared in the March 1986 issue of Harvard Business Review.

What are the problems of just

Which of these is not a goal of just

Which of the following are not the characteristics of the JIT philosophy?

What is a just