Lean Supply Chain and Logistics Management, page 7
If you recall the difference between push versus pull production, you can better understand this topic. In push, you produce in large quantities to spread your fixed costs among a large number of items, thus minimizing your costs per unit. In pull, you schedule closer to what the customer actually wants (i.e., make what you sell). The ultimate goal is one-piece flow. While this may be unattainable, it is the direction that you want, and need, to head toward.
There are many benefits to this JIT approach. In Fig. 6.1, which compares the two approaches, you will notice a few things. First, you can see that smaller batches reduce the overall cycle time for any one item. In the push approach, you have to wait for the large batches of other products scheduled before you even get to the one you are waiting for (item A in this example). You can also see that WIP is significantly reduced by the small lot approach. Finally, in the event that there is a quality problem that might affect the entire batch, it becomes less of a problem because of the batch size reduction.
Figure 6.1 Push versus pull.
Batch Size Reduction
The benefits of batch size reduction can include reduced lead times, lower inventory levels, more flexibility to meet fluctuating demand, better quality with reduced scrap and rework, less floor space used in production and storage, and thus lower overall costs.
In supply chain and logistics, there are the obvious results of batching in production to cover manufacturing wastes resulting in excess inventory, and in purchasing to obtain economies of scale (i.e., to get better pricing, which will be discussed more in Chap. 7), but it can also be seen in the office where batching typically occurs in the form of paper which can pile up in people’s inboxes. There is a natural tendency to batch in an office as there usually some kind of setup for each type of activity, such as order processing, where files, forms, faxes, and reference materials are gathered before going to a specific computer screen. Each step in the process is typically done in batches and therefore ends up sitting in someone’s inbox until they can get to it.
Quick Changeover
The primary obstacle to reduced batch size is changeover time and costs (note: the typical definition of a changeover in manufacturing is “the time from last good part to first good part”). The goal is to minimize changeover time and cost, so that smaller batches are run more frequently, resulting in better flow.
To go from one activity to another, whether on the shop floor, office or distribution center, requires some kind of changeover, which includes some kind of setup. As a result, batching seems like the most efficient way to do things (i.e., large batches equal fewer changeovers). Therefore, high setup costs encourage large lot sizes, and reducing setup costs reduces lot size and reduces average inventory.
If we can reduce the time and cost to changeover, batch size reduction can be realized. There is a concept called “single minute exchange of dies (SMED) which, while it literally applies to production operations involving a die, is used generally to refer to quick changeovers (or setup reduction), which result in smaller lot sizes and improved flow. Changeovers can vary in time from minutes to hours in manufacturing, but the idea is, through team-based continuous improvement, to keep reducing changeover time and cost so that things are produced in smaller batches or lot sizes.
Think of the race car pit crew and all that they do in a very short amount of time (Fig. 6.2). While the car is being serviced, it is not in the race, so the quicker the crew can get it back on the track, the better chance their driver has of winning the race. The same thing applies to changeovers and setups in business. If a piece of equipment is down, you are not able to make the product. The longer and more costly the changeover is, the less you want to do it, with the end result of large lot size batches. So if you can reduce the time and expense of setups, you can then changeover more often, making smaller batches and getting closer to one-piece flow.
Figure 6.2 Car pit crew.
Typical changeover tasks can involve preparation and adjustments, removing and mounting, measurements, settings and calibration, and trial runs and adjustments. If, as a team, you focus on waste in the current changeover process, it is not that hard to reduce the time it takes to changeover. Even a few minutes may be critical.
Most people involved in setups and changeovers feel that they are doing it the best possible way. However, the benefit of doing it in a kaizen event, with a mix of people from various functions, can shed new light on the subject. Typically, the team will observe and document the process (more than once if necessary) from beginning to end.
A useful tool in quick changeover is a setup analysis chart (Fig. 6.3). While a changeover is being observed (or videotaped as it may take hours), every step in the process should be documented including how long it takes, whether the step is “internal” (preparation while equipment is down) or “external” (while running) and distance traveled.
Figure 6.3 Setup analysis chart.
Some general keys to improving a setup are:
Try to separate preparation or external setup from actual or internal setup and move as much as possible to external so you can shorten the changeover time.
Move material and tools needed for the changeover closer to the actual spot it is needed.
Standardize the actual process (and combine steps where possible) and tools used.
Train operators and mechanics on procedures.
The net result should be an improved, shortened changeover.
In supply chain and logistics, as mentioned before, there is a large amount of batching of paperwork in an office, which if reduced can encourage improved flow and getting orders out faster, resulting in a shorter order-to-cash cycle.
In warehouse operations, there are setups everywhere, including receiving, picking, staging, loading, and shipping (especially for shift startups). Usually the first half-hour or more is fairly unproductive, so the more that can be standardized and visualized, the more productive personnel can be right from the start (especially temporary labor).
Figure 6.4 shows a receiving door that has been organized and visualized so that the operator can start up quickly as everything the operator needs to do the job is in place and ready to start unloading.
Figure 6.4 Organized receiving door.
Kanbans
A key to successfully going from a push to a pull environment is the use of kanbans.
A pull system typically uses signals to request production and delivery from upstream stations (it might be a card with replenishment information, or as simple as a line on the wall; for a simple example, see Fig. 6.5). Upstream stations only produce or replenish when signaled.
Figure 6.5 Simple kanban.
The tool to execute this process is called a kanban, which is essentially a way to control the flow of materials and other resources by linking functions with visual controls. Only what has been consumed at the demand by the downstream customer is replaced. This determines the production and replenishment schedule.
By pulling material in small lots, inventory cushions are removed, exposing problems and emphasizing continual improvement
There are many benefits to a pull system, including reduced cycle time, less orders “dumped” downstream creating excess WIP, less reliance on a forecast, and short lead times for customized products or services.
The most common thing you may hear about kanbans that failed is that “we had one but it stopped working.” One of the main reasons for this is the establishment and maintenance of reorder points and reorder quantities. Lead time and demand rate to replenish can determine “when” to order and the Economic Order Quantity (EOQ) model, which minimizes inventory holding and ordering costs, can be a way of determining “how much” to order. In many cases, an item’s demand may have “peaks and valleys.” As a result, it is necessary to continually review kanban order points and order sizes, as they may need to be adjusted based upon the current demand rate.
Many companies also try to use a “one size fits all” type of model when establishing kanban reorder points and quantities. This of course dooms the program to failure as each item has its own individual demand characteristics and needs to be looked at individually.
In supply chain and logistics, kanbans have many applications. Starting with the obvious replenishment of raw materials with vendors. This replenishment method can be initiated manually where a kanban card is pulled when inventory is running low. The kanban card has basic item information such as reorder quantity, pricing, etc. needed to place a reorder manually or electronically based upon a predetermined reorder point. The extreme of this is called vendor-managed inventory (VMI) (which will be discussed in more detail in Chap. 13), where the supplier takes full control over replenishing materials, saving the customer the costs of monitoring inventory and placing orders. Typically, the vendor either makes regular visits to check on inventory levels, or the information is transmitted electronically to the vendor (there are now even vending-type machines to dispense and automatically reorder parts and hardware directly from the vendor). This results in orders being pulled in small lots, more frequently than was the case before the VMI (and also eliminates the need for you to check inventory and place purchase orders).
A kanban can also be used to replenish supplies and packaging materials. A good use for kanbans in distribution centers is to replenish supplies, such as labels, tape, and corrugated boxes in an office or work area. It can be as simple as a line on a wall to determine reorder point (see Fig. 6.6).
Figure 6.6 Simple kanban.
Kanbans help to enable the concept of point-of-use storage (POUS), which is the idea of having things you use more often closer to you (and typically at waist level) and the things you use less often, further away and higher up. The kanban can be used to keep minimal amounts of raw materials and supplies nearby, without taking up a lot of space in the work area where space is at a premium and efficiency is everything. It also simplifies physical inventory tracking, storage, and handling and is a foolproof way to ensure that an area never runs out of needed materials or supplies.
Quality at the Source
Another Lean concept is quality at the source (also known as source control). The idea here is that the next step in any process is the customer, and you want to make sure that you deliver perfect products to that customer. A way to help ensure this is through the use of a poka yoke, a Japanese term for mistake-proofing. A poka yoke is a way to make so that it is virtually impossible to pass on a defective part or piece of information from one process to another (think of the saying “you can’t put a square peg in a round hole”).
This does not just pertain to physical product, where you can create a foolproof device to ensure that the correct product is passed on (e.g., creating a right-angle “jig” to test the product before passing it on). It can be used for information by limiting the choices on a form or screen, for example.
Quality at the source is typically used in conjunction with a total quality management (TQM) program. TQM is similar to Lean Enterprise in that it is a team-based program spanning the entire organization, from supplier to customer, and requires a commitment by management to have a long-term, companywide initiative toward quality in all aspects of products and services as defined by the customer.
There are seven tools of TQM. They are: continuous improvement, Six Sigma, employee empowerment, benchmarking, Just in Time (JIT), Taguchi concepts (specific statistical methods developed to improve quality), and knowledge of TQM tools such as Pareto charts and cause-and-effect or fishbone diagrams. Some of the seven tools of TQM are also found in Lean thinking (continuous improvement, empowerment, and JIT).
Work Cells
Another powerful tool that can significantly impact the efficiency of the workplace is a work cell. Work cells rearrange people and equipment that would typically be located in various departments into one group so that they can focus on making a single product or providing a single service or a group of related items or services. It is not a new idea, as it was originated in 1925 by R. E. Flanders.
Work cells are typically layed out in a horseshoe-type shape, which allows for a more efficient use of work space and equipment and is conducive to one-piece flow and, as a result, less WIP. There is also the benefit of needing less workers, as each worker in the cell is able to do all of the activities required to produce or assemble the product or to deliver the service. Employees in a work cell typically have higher morale as a result of greater participation in the entire process.
The typical first step is to identify families of products or services (a “family” should have most, but not necessarily all, of the same steps). As a result of the job enlargement, a typical feature in a work cell where employees have multiple responsiblities is that there is a great deal of training required, which results in a lot of flexibility in the cell. There is often an opportunity to use poka yokes to ensure good quality as well.
Balancing a Work Cell
It is then important to calculate the takt time for the product or service family (total work time available/units required) in order to balance the activities in the work cell so that materials or information can flow. So, for example, if the daily demand for a product (on an 8-hour shift assuming no breaks in this example) is 800 units, the takt time is 100 units/hour or 1 unit every 36 seconds. Each activity in the cell should be capable of making and passing 1 unit every 36 seconds in order to be balanced.
This information can help to identify bottlenecks in the process where one or more components or resources limit the capacity of an entire system, so as to make sure the cell is balanced properly (i.e., that each step in the cell is capable of processing one unit at the required rate of takt time). In the event that it is not balanced, having flexible, cross-trained employees can help address bottlenecks; in the case of machine bottlenecks, other approaches may be necessary, such as running overtime, speeding up, or adding equipment.
A time observation form (Fig. 6.7) is a useful tool when looking to staff and balance any multiple-step process. You take multiple measurements of the time to do each activity to see if the entire process is balanced and if there are any bottlenecks (versus the takt time). It can also help to see if you have the proper number of people involved in the process (total observed time to complete all activities for one part divided by takt time).
Figure 6.7 Time observation form.
Work cells can be appropriate not only on the shop floor, but in the office and warehouse. For example, in the office for a distributor, there are various functions that are required in order processing, including receive order, check credit, review and enter order, reconcile and confirm order, and finalize and release order. Typically these activities are done by different people, who are possibly in different departments. This activity can take a day or longer, while the actual value-added activities may only take 30 minutes or so. Along the way, there may be batching of orders, waiting for approvals, and a lot of walking around.
If this function were to be set up in one work cell performing all of the activities, the entire process could now take less than 30 minutes for each order using fewer employees (who have had their jobs expanded). The end result is that the orders are released for picking and shipment much quicker, thus speeding up the order-to-cash cycle for the business.
If you are implementing a work cell, you should realize that there may be higher wages involved as a result of more responsibities as well as more training, but it can be well worth it overall.
In the distribution center itself, there may be more limited opportunities, but they are typically found in areas like packaging or value-added activities performed by 3PLs such as packaging of kits for a customer.
Total Productive Maintenance
The final major concept in Lean that I would like to cover is that of total productive maintenance (TPM), which focuses on equipment-related waste. Equipment maintenance (or lack thereof) is often an overlooked area of waste. In fact, studies have shown that most manufacturers (70 percent or so) operate under what is commonly called breakdown maintenance. You would not treat your car like that. Every 3,000 miles or so, you take your car in to change the oil and air filter, lubricate various parts, check fluid levels, etc. This is called preventative maintenance (PM). You do not wait until the transmission drops out on the road before checking the fluid, gears, etc.
There is another type of maintenance called predictive maintenance, but that is typically used after a good PM program is in place. In predictive maintenance, tools are used to check temperature, vibrations, etc. to see if some corrective action is necessary.
TPM, in and of itself, is not a PM program. It can, however, result in putting such a program in place.
Overall Equipment Effectiveness
Basically, a piece of equipment is observed (and possibly videotaped) during an entire shift to come up with an overall equipment effectiveness or OEE percentage (Fig. 6.8). According the various studies, typical companies average in the 70 percent area. That means that there is room in most companies to reduce or eliminate equipment-related waste to increase throughput and quality.
Figure 6.8 OEE calculation.
