Production scheduling requires close coordination between sales and plant operations. In Toyota Learning, we will explain how the production schedule is used to provide consistent and continuous flow of materials and vehicles throughout the supply chain. In Sales and Operations Planning we described how information is gathered, both top-down and bottom-up, to create a three-month order and forecast. Once a three-month order and forecast are received from the sales divisions, the next step is to create a production schedule for the assembly plants. Before a production schedule can be determined, we need to understand how vehicles flow through the assembly plant.
Assembly Plant Operations
A typical Toyota assembly plant is highly integrated. Toyota assembly plant process shows how a vehicle flows through a Toyota plant.
Under one assembly plant roof you will find all of the following:
Stamping shop. The stamping shop is where the body parts for a vehicle are stamped out of huge coils of steel. Large stamping presses are arranged in a row connected by conveyors. At the start of one row of presses, sheets of steel are cut from the coil and fed into the first press. After the first press stamps the basic shape of the body part (for example, the hood or the door panel), it is routed to the next press, where the press may create a curved shape. From there, it is passed to a press that punches holes, until finally the finished part rolls off of the press line. Once a press line is configured to stamp a specific body part, it will run for a period of time until an adequate quantity of parts is produced. Typically, the inventory of stamped parts to supply the body shop will be enough to generate only a few hours of production. The reason that the stamping line is run in batches is because dies must be changed within the presses to reconfigure the line to stamp a different part. But because Toyota plants can change dies in a few minutes, several different parts can be stamped on each press line each day. Many writers have pointed out the practicality of Toyota’s practice of running batches through this operation, including Monden.1 Toyota forsakes the one-for-one ideal with the batch operation, when appropriate. However, unlike many batch operations, very little is left to randomness. For example, typical press operations might waste several blanks before getting the first part right, which is counted as setup time and cost. In Toyota, that waste is avoided by making sure that the presses are maintained and function at a high level of precision. The next step in the process is the body shop.
Body shop. The body shop is where a vehicle is born—where it gets its identity. At body start, each vehicle gets a unique body number and is assigned a Vehicle Identification Number (VIN). The vehicles are started in the planned line-off sequence, which is based on the production schedule and sequence. From this point on, each vehicle is processed individually throughout the production process. The body number is used to track the vehicle and look up the vehicle specifications so that the exact body style is welded together to form the completed body shell. The body shop is highly automated; most of the work is performed by robots that are programmed to weld each of the vehicle panels together to form the body shell. Toyota has made particular innovations in the body shop by using robots with the flexibility to build different types of body parts. From the body shop, the vehicle shell is moved by conveyor into the paint shop. Improvements in robot flexibility have enabled an increase in the body shop process flow at Toyota. Note that at Toyota plants, there is no body shell bank or in-process inventory between the body shop and the paint shop. That is another innovation based on building vehicles one by one in a planned sequence.
Paint shop. Color is applied to the vehicle at the paint shop. The shop consists of multiple paint booths that use robots to spray-paint the body shells coming from the body shop. As the vehicle body shells enter the paint shop, they are dipped in a solution that covers the body with a protective coating. Then they are organized by color and sent to one of the paint booths. Because of environmental considerations, vehicles are painted in small batches of the same color. That approach reduces the amount of pollutants dispersed into the air, as it limits the number of times the paint nozzles must be flushed (which happens each time the color is changed). Also, some colors will require multiple coats, which will result in the vehicles getting out of the planned line-off sequence. Therefore, vehicles are moved to a paint bank prior to going into the assembly shop. A radio frequency tag is affixed to the vehicle in the paint shop with the unique body number encoded. That tag enables each vehicle to be tracked throughout the assembly process, which is important because each vehicle’s specifications are used to identify parts that need to be installed on the vehicle as it moves along the line. The tag is attached at this step because it would be damaged by the paint ovens if it were affixed prior to the paint shop. The next step is to select vehicles from the paint bank to begin assembly.
Assembly shop . The assembly shop is the most labor-intensive shop. Most of the parts are installed by hand by team members working in small teams. Each team is responsible for the work performed during one process cycle at one workstation. The vehicles thus move from one station to the next at takt time intervals. Recall that takt time is the time it takes one vehicle to be completed or lined off the assembly line. At the assembly start position, a team member selects vehicles from the paint bank to start in the assembly shop. Although a computer is used to suggest the start sequence, a team member actually makes the final decision on which vehicle to start next. There are several objectives that must be considered in making this decision, among which are the following:
- Keep vehicles in the original planned line-off sequence. Each vehicle has a planned line-off date and time stamp that can be used to pick the oldest first.
- Avoid starting vehicles back to back that have high workload impact on assembly team members. For example, sunroofs may require extra work in one or more processes.
- Maintain the ratio of models on the assembly line. For example, if the line is producing two models and the ratio is 45 percent and 55 percent, then this mix should be maintained.
As the vehicles move down the assembly line, the team members receive their instructions (on manifests) about which parts to install on each vehicle. These manifests are generated by the assembly line control system via a scanner that reads the radio frequency tag. The approximate number of steps in assembling a car is 353. Out of these steps, fewer than 10 use parts from sequence suppliers. In addition, sequence parts suppliers receive an electronic transmission that advises them of the exact lineoff sequence. The information enables them to build the parts (e.g., seats) based on the exact specifications of each vehicle and ship them in the exact sequence to be installed on the line. Although, most sequenced parts are provided by external suppliers, some sequenced parts, such as plastic shop parts, are produced in-house.
Plastic shop. Although the plastic shop is not part of the in-line process of the assembly plant, it provides key components that must be synchronized with the assembly process, and it functions as an internal sequenced supplier. An injection molding process is used to create plastic parts such as instrument panels and bumpers. Because these parts are colored, they need to be sequenced to match the vehicle colors. As happens with sequenced suppliers, these plastic parts arrive at the assembly line on dollies in the exact sequence of the vehicles in which they are to be installed.
Now to return to the production line direct flow and examine the inspection process:
Inspection. After the vehicle is completed and comes off the main assembly line, it is driven to the inspection line. The primary purpose of this inspection is to conduct functionality tests on such components as the engine, transmission, brakes, and air bags. Next, cars are subject to water-pressure testing to ensure that the vehicle is leak proof. All teams involved in the assembly process are instructed to stop the line in the event a defect is detected or a problem identified. Thus it is not necessary for Toyota to perform rigorous inspection after a vehicle is completed. If a vehicle is identified as having a defect, then it will be diverted to a repair area so that the defect can be corrected before it leaves the plant. The final step before the vehicle leaves the plant is the sales line.
Sales line. The point at which a vehicle’s ownership is transferred from the manufacturing division to the sales division is the sales line. As the vehicle passes an imaginary point on the line, the manufacturing team member scans the vehicle into the plant computer system to change the status to “sold,” and the sales team member scans the vehicle into the sales computer to change the status to “bought.” That point is also where many of the so-called throw-ins are placed in the car trunk or glove box; those items would include floor mats, wheel covers, and manuals. Once the vehicle is lined off and bought by the sales division, the assembly process is complete. The total time it takes a vehicle to move from body start to line-off obviously will vary by plant. However, it usually takes about two calendar days, or three to four production shifts.
Another way to grasp an understanding of the production process is to take a tour of one of the Toyota plants. During the writing of this book, Ananth Iyer and Roy Vasher went on a guided tour of the Toyota Motor Manufacturing Kentucky (TMMK) plant in Georgetown, Kentucky. Ananth Iyer captured the following notes during the tour:
The Toyota factory at Georgetown, Kentucky, covers over 1,300 acres of land and employs about 7,000 team members. There are another 1,500 to 2,000 people employed by vendors working on-site, such as food service, day-care workers, etc.
The starting point at the plant is coils of steel delivered about every 30 minutes, each coil weighs up to 24 tons. Toyota has 19 press lines and 33 presses. Eight-hundred-ton presses operating at 80 strokes per minute create blanks. An example of how Toyota is continuing to kaizen its operations is that a new stamping line is being installed that is expected to save over 32 percent in energy costs and replace two existing lines.
The blanks are transferred by robots to a press that converts parts to requisite shapes. There are over 1,300 dies on site. These components are transferred to a flow rack and conveyed to body weld. The body weld operation takes 274 sheet metal parts and welds them to form a shell. There are over 700 robots that perform more than 4,400 welds to create a shell. Team members rotate tasks every two hours so that they use a different muscle group for their work and reduce monotony.
Of the 20 hours it takes to make a car, around 9 hours are spent in the paint shop. Every 55 seconds of takt time a car is completed. The plants at Georgetown produce nearly 2,000 vehicles every day. Production in the line is in a mixed sequence and varies by color, across Camry, Solara, and Avalon. The paint shop has 20 different colors of paint, but the most popular colors in September 2008 were silver for the Camry and bright red for the Solara. In addition, one in every four Camry models produced is a hybrid.
When the car leaves the paint shop and arrives for assembly, sequenced suppliers receive notification to deliver their parts in the exact sequence that cars are produced. In addition, the doors are removed to permit easy access and to prevent damage to the doors. The door components are assembled separately. The exact door of a body rejoins the car at the end of the line. Assembly line associates operate in teams and use the andon cord (an andon is similar to the cord on a train that when pulled sends a signal) to shut down the line when a problem is detected. The team leader helps fix problems immediately, but if a problem is not fixed within a cycle time, the line is stopped.
The TMMK plant has andon cords pulled over 5,000 times each day. Each area has a different song pattern. Our trip itself saw many different line interruptions. The immediate attention to problems guarantees that quality is built in during production for every car. But it also means that adhering to the production plan is difficult. The role of inventory at the paint shop and at other points along the line is to enable the system to recover from disruptions that may change the assembly sequence.
All along the plant there were “blue walls” with information regarding the daily production, productivity, and so on. The data enable managers to “walk the wall” and get a quick read of the plant’s operating performance.
The plant and its carefully planned and deliberate pace and methodical execution of tasks have provided a glimpse of a microcosm of the Toyota supply chain in operation.
Now to examine how the production schedule and sequence are created. Some of the metrics to monitor production are first-run ratio (the percentage of vehicles that go through the line and are completed on the first pass without being pulled off the line), actual sequence versus planned sequence, and actual line-off time versus planned line-off time.
Production Scheduling
The production schedule is created once a month from the sales order and forecast. As discussed in Sales and Operations Planning, sales divisions submit a rolling three-month order and forecast each month. The Production Control division must create a daily production plan to execute the agreed-upon schedule.
Scheduling Inputs
For the purpose of this discussion, assume that the next production month “N” is being scheduled. The “N” month is considered a firm order commitment, and “N 1” and “N 2” are considered preliminary forecasts. The difference between the firm orders versus the forecast is that once the firm order is submitted from sales to manufacturing, the volume of vehicles by model by plant is frozen. In other words, sales divisions have committed to buying these units, and manufacturing has agreed to produce them. Nevertheless, the content of the vehicle specifications can be changed up to about a week prior to line-off.
The forecast for months N 1 and N 2 do not prevent changes in volume or content. However, because of the 80/20 rule of mix planning described in Toyota Mix Planning, the actual variation in individual options from month to month will be somewhat muted when calculated on a daily rate basis.
Another input necessary to create the production schedule is the production calendar and the operations plan for each plant. As discussed in Sales and Operations Planning, the production calendar consists of four or five weeks for each month; however, the holiday schedule will vary by plant. For example, Thanksgiving is celebrated in Canada during October and during November in the United States. Each plant also uses a different operating plan that quantifies the number of vehicles to be built on each production day. The quantity per day may even vary by day of the week. For example, planned overtime may be two hours per day from Monday to Thursday and zero hours on Friday. Such flexibility demonstrates that Toyota makes extra effort to accommodate the quality of life of team members.
The final input that is required is the constraints. Those constraints could be a limit on the type of engine that can be built on one of the assembly lines or that certain colors can be built only on one line. Another constraint could be a ramp-up or ramp-down of a specific option. That occurs when a new option or color is introduced as a running change in midmonth. The constraints are established by each plant each month and reviewed by the production planners to ascertain that they are necessary and reasonable. Feasible production schedules have to satisfy these constraints.
Scheduling Process
The first step in creating the production schedule is to use the sales order and forecast data to create individual records for each vehicle and assign a unique reference number. That step is necessary because each vehicle has to be assigned to a production slot. Though eventually the Vehicle Identification Number (VIN) can be used to identify a unique vehicle, the VIN is not assigned until the vehicle is started in the body shop. Therefore, a Unique Reference Number (URN) is assigned to identify vehicles prior to production line-off.
A heijunka process is used to schedule the vehicles by day, by line, by plant. Heijunka (or smoothing) is a technique to avoid supply chain congestion, workload imbalance, inventory batching, and the like. The software that includes the heijunka logic is proprietary, so the details will not be described. However, the concept of heijunka is to create a level, or smooth, production plan. This concept of heijunka is also called “mixed-model production.” The benefit of heijunka within the plant is to smooth capacity requirements and balance use of resources. The concept extends beyond the shop floor. By smoothing the flow of dependent parts, Toyota makes sure that its parts suppliers also see a level load. In fact, it ensures a level load for parts even from Japan or from distant suppliers by restricting the day-to-day variation to between 5 percent and –5 percent of the supplier’s order.
Usually some sales orders have specific build dates requested, so those orders need to be scheduled first. One such example is fleet orders. The large rental companies such as Hertz, Avis, and National require that their monthly shipments arrive during a specified time period each month. Because of the limited space at most of their rental locations, they attempt to stagger their new vehicle arrivals and the shipments of the used vehicles. Therefore, these orders need to be scheduled based on a date range such as the first week of a month. Another example is an individual special customer order that needs to be prioritized. Such orders are usually scheduled early in the month so the customer will receive his or her vehicle as soon as possible.
Next, the remaining orders are grouped by build combinations and spread throughout the month so that the number of identical orders will be evenly distributed across the month. Then the sum of each option for each day is checked against the constraints. That will result in a need to shuffle some orders around to ascertain whether the constraints are met. As you can imagine, doing so is like trying to solve a Rubik’s Cube, because as you move one type of order to resolve a constraint on one option, it will create a constraint violation of another option. To avoid an endless loop when trying to obtain the perfect heijunka for each option, a priority weighting is assigned to each option to determine its ranking. Priority weighting is similar to rate-based planning; for example, demand can be imagined as a rate, production as a rate, and supply as a rate. Constraints on capacity are limitations on rates of different important supplies. If those rates do not match, then there will be creation of inventories or back orders.
In addition to options, the destination of vehicles is also considered as one of the heijunka factors because it is important to have an even flow of vehicles to each region. Thus, the analogy of rates is carried forward to rates in different directions. The goal of heijunka is to balance these rates.
Once the heijunka process is completed, then each order is assigned the scheduled production day. The production day is deemed to mean the scheduled line-off day (i.e., the day that production of the vehicle is completed). This production schedule is then sent back to the sales division to advise its members of each vehicle’s Unique Reference Number. The production schedule is also sent to each plant to make the actual production sequence.
A metric to measure the stability of the production plan is to measure the smoothness of the heijunka by option. Production Sequence Each plant must determine the exact production sequence within each production day. The sequence is determined by the operational conditions within each plant. Some of these conditions are color batching, workload associated with specific labor-intensive options, and heijunka of the major options within a day. Similar to the heijunka logic, this logic is proprietary and is considered a black box. The daily production plan is the input, and the output is a production sequence for each day. This production sequence is used by the plant to create the parts orders. It is also used as input to the assembly line control system so that vehicles are started in the correct sequence.
Sample Production Plan
Assume that Sample Set of Vehicles contains a list of vehicles that are to be scheduled to create a production plan. For this simple exercise, there are only three options for each vehicle (i.e., grade, engine cylinder, and color). The grade can be either LE or XLE. Engine is either a fouror six-cylinder. Color has three choices: red, black, and blue.
The objective is to create a production schedule for these 10 vehicles over a five-day period and to achieve a level quantity of each option. The ideal schedule would contain an equal number of each grade, engine, and color per day. But as you can see, achieving that would be impossible because the sum of each option is not divisible by 5. For example, there are six LE grades and four XLE grades.
Scheduling Template is a sample template that illustrates how the schedule results will be shown by option by day.
The first step is to group the vehicles by unique build combinations. As you can see from Grouplike Build Combinations, there are two vehicles that are in group A. They both are LE, four cylinders, and blue. In this example there are seven groups. It is important to identify how many vehicles have the same build combination, because if you spread the groups across the production days, you will automatically smooth several options.
The next step is to sort the vehicles from the most important priority to the least important priority. Priority is assigned to the specifications that are most important to the plant production flow. Doing so will enable the scheduling process to start with the first vehicle and schedule the vehicles in sequence day by day. In our example, the highest priority is “grade,” followed by “engine.” Vehicles Sorted by Highest Priority Specification shows the result of this sorting process.
Results of First Scheduling Pass shows the result of scheduling vehicles based on the most important option: grade. That schedule is obtained by distributing the orders from the list uniformly across the days of the week. As the table reveals, the result is not perfect, because there are six LE grades to be scheduled over a five-day period. When the number is uneven, then the goal would be to make the best fit. In that case, there will be two LE vehicles scheduled on the first day. Now if we look at the second-priority option (engine), there is an uneven schedule on days 2, 3, and 5. On days 2 and 3, there are two 4-cylinder engines scheduled on both days. On day 5, there are two 6-cylinder engines scheduled.
The next step is to attempt to rebalance the vehicles based on making a smoother engine distribution without breaking the smoothness of the grade. In our example, that could be accomplished by swapping vehicles E10 and G2. The results of the second pass are shown in Results of Second Scheduling Pass.
The final step is to create the production schedule for each vehicle by assigning the production day and production slot sequence to each vehicle. Final Scheduling Sequence shows the final schedule of each vehicle. The output is then used by the sales division to allocate vehicles to dealers. In addition, the output is used by the assembly plants to create the parts order.
In our simple example, it has been fairly easy to manipulate the vehicles to arrive at a smooth schedule for all options. That task is much more complex when there are thousands of vehicles to schedule with hundreds of build combinations.
Why Is Heijunka Important?
As illustrated previously, heijunka is one of the foundational processes that has enabled Toyota’s extended supply chain to operate as if it is an extension of the TPS. There have been many books written about TPS and how it is synonymous with “lean production.” What Toyota does by establishing a smooth production schedule using heijunka is to ensure that its own assembly plants are operating in an efficient and effective manner while at the same time extending stability throughout the supply chain.
Toyota understands that the cost of sales includes the total cost of operating the supply chain, not just Toyota’s internal production costs. Let’s consider how heijunka can positively impact all elements of the supply chain, including the various tiers of suppliers, inbound logistics, assembly operations, and outbound logistics, as well as dealers.
Multiple tiers of suppliers exist (this tier structure will be discussed in greater detail in Managing Suppliers). Tier 1 suppliers receive their orders directly from the OEM and are responsible for producing parts based on the pickup schedule provided by Toyota. Tier 2 suppliers are the tier 1’s direct suppliers and receive their orders from the tier 1 suppliers; they must produce parts or materials based on the tier 1 supplier schedule. This process continues backward throughout the network of suppliers. Now assume that the orders from the OEM have not been level for each daily order. For example, on day 1, the tier 1 supplier received an order for 1,000 parts; on day 2, the order was 500; and on day 3, the order was 2,500. Next, the tier 1 supplier broke down these orders into their component parts and sent the order to the tier 2 supplier. In this example, if there were four parts per order, then the tier 2 supplier would receive an order of 4,000, 2,000, and 10,000. Now let’s assume that the tier 1 supplier’s daily production capacity is 1,000 and the tier 2 supplier’s daily capacity is 5,000. One of two things could happen: either each supplier would stock extra inventory to enable it to fulfill the demand or it would ship short and create a back order until it could catch up. That would create a “bullwhip effect” (as examined in The Beer Game and the Toyota Supply Chain) and lead to inefficient operations at all tiers of suppliers, especially if these variations could not be forecasted in advance. Thus, even with the best intentions, the variance of orders would exceed the variance of demand (the bullwhip effect) unless efforts were made to dampen the effect. Even though suppliers adjust and attempt to respond to these dramatic changes in demand, the inefficiencies result in higher operating costs, which are passed along to the OEM.
Now let’s consider Toyota’s orders based on the production schedule using heijunka. As we discussed earlier in Toyota Learning Principles and the v4L Framework, it is not possible to create a perfect heijunka for all parts; however, a good heijunka result would be a variation of between 5 percent and -5 percent. If Toyota’s average parts order were 1,000, then the expected daily order to the tier 1 supplier would be between 950 and 1,050 parts. Therefore, the tier 1 order to the tier 2 supplier, based on four parts per order, would range from 3,800 to 4,200 parts. In that case, because the variation in daily orders would be very small, the suppliers could adjust their daily production by varying the level of overtime instead of maintaining high levels of safety stock or risk shipping short. The next segment of the supply chain is inbound logistics. Heijunka plays an important role in smoothing the daily shipments from tier 1 suppliers to the Toyota assembly plants. Toyota uses third-party logistics partners to manage and operate a fleet of trucks that picks up parts on a daily basis and delivers them to multiple assembly plants. Heijunka of the parts volume for each supplier in the network ensures a consistent flow of parts through the logistics network. That maximizes transportation efficiency by facilitating a high utilization of trucks and drivers every day.
Once the parts arrive at the assembly plants, they come under the control of the internal logistics group. Internal logistics is responsible for moving parts from the dock to the line side just-in-time. Kanbans are used to signal when the parts are needed for each line side station. Again, heijunka ensures a smooth flow of parts within the plant. That enables the forklift drivers to operate in an orderly manner on a regular internal route schedule.
Once the vehicles are produced, they are ready for shipment to dealers. Vehicles are transported by various methods including ship, rail, and truck. In the United States, however, most vehicles are transported by rail to a regional railhead. After the vehicles arrive at a railhead and are unloaded, a trucking company picks them up and delivers them to the dealers. As discussed earlier, Toyota strives to ensure that all components of the supply chain are streamlined and operate efficiently and effectively. Therefore, if vehicles are produced and shipped randomly without regard to railhead destination, then vehicles arrive in an uneven manner and eventually create a bottleneck. To prevent that from happening, Toyota includes the destination code as one of the parameters that is considered in the heijunka process. That inclusion ensures a smooth and even flow of vehicles through the distribution network.
Finally, the vehicles arrive at a dealer and are placed into inventory until sold. (Remember driving past dealerships that proudly park their vehicles in large lots in front and around the dealership? Those vehicles are the inventory.) Again, you may question, how does heijunka affect the dealerships? Just as at the railhead, a bottleneck can occur at a dealer if too many vehicles arrive in a short period of time. Dealership personnel must prep the vehicles once they arrive and get them ready for sale. It is best if this work is spread throughout the month. In addition, it is important to have a steady flow of vehicles to each dealer to avoid unnecessary buildup of inventory. Heijunka is again used to smooth the number of vehicles scheduled throughout the month by region. The region then allocates its vehicles to the dealers proportionally based on sales volume. Doing so will ensure a smooth flow of vehicles to each dealer throughout the month.
In summary, heijunka plays a vital role in Toyota’s supply chain operations. It is used to create a smooth flow of parts from suppliers to the assembly plant as well as for maintaining a smooth flow of vehicles from each assembly plant to the dealers.
Why Is Production Sequence Important?
The sequence the vehicle is produced within the day is important because the assembly plant operations need to be well balanced to ensure that there are no bottlenecks within the production process or overburden on selected teams. Again, Toyota’s focus is to guarantee smooth operations throughout the assembly process. In an assembly plant, there are hundreds of workstations along the line that install parts on the vehicle as the vehicle body passes through. The objective is to make sure that each work team’s effort is similar and that the work can be completed within the takt time. Therefore, the production sequence is established based on smoothing options that create extra work for one or more work teams.
Once the sequence is set within each day, then the estimated line-off time (or completion time) can be assigned to each vehicle. A typical production day for a plant that runs two production shifts starts at 6 a.m. and completes at about 2 a.m. the next calendar day. So the first vehicle would have a planned line-off date/time of production D: 06:01, whereas the last vehicle planned lineoff date/time would be D 1: 02:00.
The planned line-off date/time is used by sales to calculate the estimated time of arrival (ETA). The ETA is used by dealers to keep customers posted on the scheduled arrival of their vehicle. The process to create the ETA will be explained in Logistics.
The line-off date/time is also sent to the parts ordering process to determine when parts will need to be shipped from the suppliers to arrive just-in-time to be installed on each vehicle. This process will be discussed in more detail in Parts Ordering.
How Does Toyota’s Scheduling Process Compare to Others?
Master production scheduling (MPS) is the process used in the manufacturing planning and control framework to initiate more production. A master production scheduling process plans production as forecasts are updated and also when orders are received. The master production schedule is often determined at the group level. [The final assembly schedule (FAS) also coordinates between the production plan and the rest of the manufacturing processes by specifying the exact build sequence.] A check is made to ascertain whether the aggregate of the detailed planned build equals the volume planned by the MPS. The master production schedule is the input to materials planning. Materials Requirements Planning (MRP) will be discussed and compared to Toyota’s method of materials planning in Parts Ordering. At this time it is enough to note that the material plan uses fixed lead times to decide when to schedule parts or assemblies.
Let us now compare these MRP processes to the scheduling processes at Toyota. MRP processes share many similarities with scheduling processes at Toyota; they have some differences too. As with scheduling at Toyota, there is an attempt to freeze the production plan using planning fences when developing the MPS. Typically, the planning horizon is split into zones that are called “ice,” “slush,” and “water.” The ice part is frozen, the slush part is where some changes at the product family level are allowed, and the water part is open to changes. However, with MRP the attempt to manage the selling to match capacity is not undertaken with as much assiduousness as in Toyota.
It is well known that MRP tends to generate a significant amount of nervousness. That is so because small perturbations to the demand can lead to significant variations within the plant as well as for suppliers. This phenomenon is inadvertent and unavoidable because of the rules used to plan production. In MRP, production is considered to be taking place in discrete time (weeks, days, or hours); thus, the quantity to be produced is often “batched” so that the batch is the right size and started at the right time. Any such attempt at batching can lead to large changes in production requirements because of a small change in production plans. For example, say an economic batch size is 50. The demand in a period is 48. The next batch may not be started until the next period if the on-hand inventory is enough to cover current consumption requirements, even if the safety stock is low. For example, if safety stock is 10, current consumption demand is 30, and on-hand inventory is 43, then the planner might decide not to release a batch for production. Now, assume that either a customer places an order for five units or there is a change in forecast by the same five units. Then it is likely that a batch of 50 is released for production immediately. Toyota avoids such a scenario by using heijunka to create a level production plan. That stability prevents the nervousness associated with rules of batching in MRP within the plant and even when ordering from suppliers.
The materials plan is traditionally executed with a manufacturing execution system (MES). Often, the traditional scheduling process has to contend with managing orders through several complex steps to achieve the lead time promised by the material planners. For example, a stamping shop might have several processes. For material planning purposes, manufacturing a particular stamping is considered to be a single process. The average lead time plus some slack is used to plan production of this part in the shop. The shop therefore has leeway in scheduling the individual orders such that they meet the lead time; safety stock to cover scrap and rework is also added. The inclusion of slack in planning and the lack of step-by-step coordination inevitably lead to carrying inventory as work-in-process or finished goods. They also lead to temporary surges in capacity requirements that are seemingly unpredictable. In Toyota’s process, the production plan is based on each shop working at its standard operations rate; thus, surges in workload are avoided. That approach makes scheduling easier to accomplish. Deviations are obvious and visible; they can be traced and addressed as and when they occur.
We say that the traditional (and common) methods use position-based planning because the position of inventories dictates the production planning and scheduling, not the rate of demand and supply. In summary, the processes mentioned previously deviate fundamentally from Toyota’s because they use position-based planning instead of rate-based planning. They also differ significantly because of the lack of a “self” or “automatic” coordination/constraining mechanism such as heijunka, which forces collaboration across production and logistics planning, scheduling, and supplier planning processes.
Reflection Points
- Variety is planned and distributed across periods (using heijunka) to balance tasks.
- Velocity is maintained using a rate-based planning of flows balanced across the supply chain. By eliminating bottlenecks, the velocity is maximized.
- Variability is curtailed with heijunka to smooth out workload and loadings. That variability reduction enables suppliers to plan their capacity reliably and thus lower costs.
- Visibility is ensured by eliminating inventories, simplifying planning, ensuring buy-in, and so on. The following are highlights of the learning practices:
- Create awareness. Heijunka makes deviations evident and forces planners to resolve issues as they arise. It enforces coordination at the supply chain level and makes problems evident to supply chain participants.
- Establish capability. Production control planners undergo intensive apprenticeship. Senior planners are asked to devote time to training.
- Make action protocols. Methods for taking actions to resolve heijunka are documented. The sequence in which different constraints are considered during planning are discussed and documented.
- Generate system-level awareness. Systemwide implications are captured by heijunka itself. It supersedes immediate concerns about local profit and loss. Heijunka makes deviations evident and forces planners to resolve issues as they arise. It enforces coordination at the supply chain level.