Read the article `JIT and MRP II could make beautiful music together’ (Resource Item 8.1). This article ties together topics we dealt with in this unit, Unit 3, Process Design, Unit 6, Quality Management, and Unit 9, Time Based Operations. The article emphasises that an integrated approach is likely to be more successful. You will benefit by reading the article again after Unit 9. In the meantime, consider whether the integrated approach described is likely to apply in all manufacturing environments.
JIT and MRP II could make beautiful music tog
Making the right choice by implementing JIT changes on top of a well-founded MRP II planning system can be the right decision.
Many manufacturing companies are aware of the potential benefits of migrating to a Just-in-Time (JIT) environment. Usually they have already spent years and hundreds of thousands of dollars implementing MRP II systems. Most people are under the impression that the two, JIT and MRP II, are diametrically opposed systems and to go to JIT would require scrapping their MRP II system. There are also those companies without formal planning systems that want to go straight to JIT. Both of these situations are a result of many misconceptions which have delayed the adoption of many of the JIT principles and processes. The strong MRP II planning environment is what will make a JIT execution environment work.
It is imperative that a company has a well-tuned planning system to assure that materials are available for the `pull scheduling’ system that JIT provides. Companies with little or no planning constants tend to overcompensate by making more product earlier than necessary, thus over-building inventories. MRP II planning systems address overcompensation by making sure that only the right materials and resources are available at the proper time. JIT espouses reduced lot sizes and lead times while making parts and products only to satisfy real demands instead of economic order quantities (EOQ) rules. Both MRP II and JIT have the same message MRP II embodies `pull planning’; making the product just before it is needed.
When JIT is implemented, it becomes critical that the master schedule be as stable as possible (that is, firm) for the next planning period. In order that this happens, the MRP II system must be functioning well with the supporting procedures and processes in place. JIT will have an effect on all areas of the manufacturing planning and control system (Figure 1). The `back end’ is changed most radically, in particular with the way the manufacturing and purchase orders are released to the plant floor Scheduling will be very dynamic, pulling the orders, while the push planning is applied to assure the materials are available. I am assuming a basic understanding of MRP II concepts in the following discussions.
Execution is the building of what the customer requests (pull) as they need it.
MRP II and JIT
Mid & long range parts ordering
Daily parts pull into manufacturing
Execution is the building of what the customer requests (pull) as they need it.
Figure 1. Manufacturing planning and control systems
It is important at this point to define what is meant by JIT and discuss its basic features.
JIT attempts to minimise all waste in manufacturing. This does not mean only in the area of inventory and scrap. Waste is anything that does not add value, and value is anything that increases usefulness and reduces cost. All areas of the business are examined to eliminate any unnecessary expenditure of time and energy. Errors are investigated to determine their cause and reduce any recurrence. It not only tries for zero inventories, but zero transactions and zero disturbances.
Some of the key features found in a JIT environment depend on a stable master schedule that balances loads and capacity. The operations must be linked and balanced, providing a smooth and timely movement from operation to operation. These operations are usually set up so they can be visually controlled by the operators and are usually executed without paperwork or complex overhead support.
Master scheduling should plan the monthly `production rate’ in lieu of discrete batch orders. The weekly/daily rates should be scheduled based on these rates. Typically, the work-in-progress inventory is reduced. Inventory is considered a waste and, as it is gradually reduced, problems are uncovered and corrected. For purchased materials, the ties to the vendors are developed to a partnership level and the delivered goods increase in quality and arrive more frequently. Product design should be determined by the customer’s definition of quality, therefore trending toward more customisation of products.
Success hinges on change
n an environment where many different product mixes are produced in any given month, reduction in the set-up time will be key. Typically, lot sizes spread the cost of expensive setups across many parts. If you can reduce the set-up times, it follows that the lot sizes can be reduced accordingly. Reductions can be accomplished by separating setups into internal and external. In 1985 Shigeo Shingo estimated that set-up can be reduced 30 to 50 per cent merely by
doing this. Internal setups are those that must be done when the equipment is not operating and external setups can be accomplished while the equipment is operating. External setups could then be accomplished while overlapping the previous operation. The internal set-up processes should be analysed and set-up methodologies put in place for consistency. Changing the layout from functional to cellular also has the effect of reducing setups since the cells are typically laid out based on group technologies. This may not be feasible throughout the entire plant but may be possible for certain component lines or operations. Replacement of older equipment with newer technology will usually have the added benefit of much reduced set-up times. Other methods for reducing set-up times are:
• Colour coding
• Standard die heights
• Locator pins
• Quick disconnects
• Duplicate tooling for smaller tools
• People involvement
• Many small incremental improvements
• Tool and die maker involvement from the beginning.
Shrinking the lot size will allow much more flexibility in the planning of what is built, how many are built and when they are built.
Another required change is a `continuous improvement’ program for higher quality or total quality management (TQM). It must be implemented from the product design, through the process design and continue through the supplier of purchased parts. Statistical process control (SPC) focuses on the operation’s critical processes. It attempts to stabilise each process by spotting trends with the goal of improving the quality of the yield. Operators must be responsible for the quality of their output and have the authority to stop production when necessary. Total productive maintenance (TPM) focuses on the maintenance of both the equipment and process, again with the goal of the highest quality output. Foolproof operations are an attempt to eliminate poor quality by examining the operations for areas where mistakes are likely and making it difficult to make that mistake again by either changing the process or providing checkpoints to catch the error before it happens. A `fishbone’ analysis should be used to determine the cause and effect of production quality problems. Preventive and predictive maintenance should be used to maintain the equipment and tools. Finally, using operator knowledge and close observation of equipment, the operator can sense potential failures. The costs of defects will be reduced as these programs are introduced, because the defects are caught before or as they occur. Continual improvement means making thousands of smaller incremental improvements and not necessarily large steps forward (Figure 2).
Figure 2: Total quality management helps make quality free
The U-cell layout
The cellular manufacturing layout, or Ucell layout, causes many additional physical changes in addition to set-up time reductions. Group technology should be used to identify parts made with similar processes, thus determining the most effective layouts. Because the machines are laid out based on the routing, the distances are obviously shorter from operation to operation, thus requiring much less handling and resulting in less damage. Since the U-cell may take the raw materials through many levels of the bill of material, there is the added benefit of fewer manufacturing orders released to produce a component. If you previously released an order for each level and returned the product to the stores area between each release, you now reduce all that travel and material handling. The workers in the U-cell layout must be capable of running several machines as well as providing basic maintenance to those machines. They can see and react to problems more quickly and provide help to each other when necessary. This flexibility must be nurtured and training must be provided to make this a reality. The operations within the U-cell should have some flexibility built in to provide for volume fluctuations. This `band width’ flexibility is especially critical at potential bottleneck stations (Figure 3). The use of value added efficiency (VAE) to measure manufacturing efficiency is key at this point. Since VAE is operations time divided by manufacturing lead time, you will get a more accurate picture of the percentage of time each part is being processed. As the VAE percentage is increased, significant lead time reductions are occurring, thus responding to customer demands much more quickly. The planning for the loads on each line should be as level as possible and balanced to reduce the fluctuations as much as possible. This should be accomplished by freezing the master schedule for that period.
How does JIT affect MRP II?
Using Figure 1 as a guide and the base knowledge of JIT just described, letís look at how a company can move toward a JIT operation and the effects this would have on the manufacturing planning and control system.
Looking first at the `front end’, let’s address the demand management area. Since a company will be making some of each product daily via a more stable, level production rate, the lead times for products will be reduced. This, in turn, will reduce the on-hand finished goods inventories. As the inventories shrink the company will slowly shift from a make-to-stock to an assemble-to-order or make-to-order company. The company will become much more responsive to customer wants and needs by supplying the right product at the right time without carrying the associated inventories. Forecasting of product families or groups will be more critical with less emphasis on forecasting individual items. It will be very important for the production plan and the master production schedule to be checked against the resources to assure level loading. The master production schedule must be rate-based and be firm for the following month or period so that weekly and daily production rates can be determined. This will result in smooth shop operations with the minor variances in volumes being handled by the use of the band widths within the cells.
Figure 3. Sample volume `band width’ for fabrication process
Looking at the `engine’, the main impact there will be in the MRP and CRP areas. Because of the redesign of the manufacturing process, either with the Ucell or actual redesign of the product, there will be far fewer component product numbers for the system and people to deal with. Reducing the number of parts also reduces the number of levels of the bill of material for that part. This drastically reduces the time and complexity of the MRP runs. Many companies have reduced MRP planning by as much as 75 per cent. Since the daily rates of the equipment would be known, as long as the production rates are within this limit, it would not be necessary to make any CRP runs. There will be significantly fewer manufacturing orders to release and, therefore, much less movement of materials into and out of stores.
Looking at the ‘back end’ you will see the more significant changes. Because of the reduced cycle time, orders move through the plant more quickly. Tracking from operation to operation becomes unnecessary and a waste of time, as the order will probably be completed before anyone gets a chance to look at the information on the order status. This eliminates the need for a complex shop floor control system. It may become necessary only to track the order’s entry and exit from the system. When the finished product is received back into stock, you back flush the component inventories to update their balances. Logistics transactions reduced or eliminated are ordering, execution, confirmation of materials moving from location to location, shipping, receiving and expediting. There will no longer be a need to create this detailed work-in-process accounting system, and the high volume of transactions associated with it will no longer be slowing down the system. This type of system requires a great amount of data integrity. This simplistic approach is the result of well designed products, manufacturing cells and systems. With the higher quality being produced, the products will flow through the process smoothly and quickly, again reducing the need for detail records and overhead staff.
Vendors must adopt JIT as well
Once JIT is working effectively in-house, a lot of the same principles can be extended to vendors. Relationships with vendors must be based on trust, with quality as the main objective. Based on the vendor’s capabilities you may be able to take advantage of the supplier’s competencies and actually include him in the product design effort. In other situations, you may have to increase the technical support you provide the vendor if the design is complex and beyond his capabilities. A prerequisite to this is the ability to provide a reasonably certain, fairly stable production schedule. Good planning in the MPS precedes excellence in execution, both internal and with vendors. Good communication will become critical, as will a vendor certification program. Electronic kanbans can be created through an EDI system, which will allow the sharing of schedules and forecasts. It will take time for the vendor to provide the quality and timeliness required, just as it will internally. Flexibility and learning together will be important while doing business. As quality improves, the number of inspections of incoming components can be reduced or eliminated. The increased frequency of deliveries will also help the vendors determine when and where they are producing poor quality. A natural result of this process will be a pruning of vendors and much more contact with those remaining, in some cases involving the line workers themselves. Using standard size kanbans and delivery to the line eliminates additional paperwork and checking. Partnerships should be developed, nurtured and kept as simple as possible.
Throughout this process, mutual respect is key between all organisations, employees, customers, vendors and shippers. A major team effort will be required that involves flexibility and authority being given to each member and using everyone’s mind and ideas, not just their hands and backs.
Case history example
A recent situation starts to bring many of these concepts home. The company is a make-to-stock manufacturer of trailers, and an MRP II user with a great amount of control over its inventory and manufacturing orders. One would think that this situation would make the company very competitive. Unfortunately, today’s environment requires more than a good planning system. Shorter cycle times in a make-to-order environment are becoming the norm.
This client makes 28 models of trailers to which can be added certain features and options. The firm makes 96 per cent of its components from the raw material level. This results in a tremendous number of orders being released to produce the 3,000 components needed for the final assembly stage. All components are kept in a storage area almost as large as the manufacturing area itself.
A tremendous amount of time is wasted in the picking and movement of materials to and from stores and the operations. There are two final assembly lines allowing each model to be built once a month. It takes the pickers three days to pick the components for final assembly. The equipment on the floor is functionally organised causing a tremendous amount of movement between functional areas or back and forth to the stockroom. Materials, labour and orders are all tracked manually on paper with updates from four to 24 hours late entering the system.
To move this Class A MRP II user towards a JIT environment, we suggested the following changes:
• Change many of the operations to Ucell layouts where feasible, thus minimising handling, movements and overall cycle time
• Change the bills of materials to reflect the cell layout, therefore reducing the number of levels in the bill and its overall complexity
• Cross train the highly specialised workers to be able to perform more operations and work different equipment. Also change the incentive system to support these changes
• Use the finished goods inventory as a base to fill orders and as individual trailers are sold, release an order to the plant to replace those trailers, thus pulling through the plant orders to replace only what is being purchased
• Implement a data collection system to track orders, materials and labour. Since orders would be moving through the plant much quicker, the inventory and orders need to be updated on a more timely basis.
This company is indicative of many companies. They are looking to stay competitive by utilising and combining the best support systems and processes available. Implementing these JIT changes on top of a well founded MRP II planning system will truly be `pulling it all together’.
Thomas Vollmann et al., Manufacturing Planning and Control Systems, Richard D. Irwin, Inc. , 1988.
Ellis, S and B. Conlon, `JIT points the way to gains in quality, cost and lead time’, APICS ¬ The Performance Advantage, August 1992.
Mirsky, M., `The Missing Link’, APICS ¬ The Performance Advantage, September 1992.
Anderson, C., `The IBM Austin CFM Story’, IBM, Austin, Texas.
Bowman, J., `Just-in-Time and MRP II: A Winning Combination’, APICS ¬ The Performance Advantage, October 1992.
Production and Inventory Management Journal, Volume 33, Number 3, American Production and Inventory Control Society, 1992.
Credit: Figure 1 from Manufacturing, Planning and Control Systems, Thomas Vollmann et al., Richard D Irwin Inc., 1988.
The author: John W Burgan
John W Burgan, CPIM, has over 20 years’ experience working with manufacturers and is currently affiliated with Skill Dynamics/IBM in Marietta, Ga.