Many distribution companies can lay claim to having system bottlenecks, especially during peak times. These bottlenecks often result in operational fire drills and system workarounds, both costly ways of doing business. Though some workarounds may seem inevitable, they should be minimized, and those that are necessary should be formalized into standard business practice (to meet business spikes). Operations that do not minimize system workarounds and that do not implement formal procedures to accommodate peaks and other deviations from standard business generally have poor productivity and accuracy as well as inflated distribution costs.
Workarounds = Chaos
Realizing that a distribution system is comprised of many unique functions, these functions must be balanced if system bottlenecks are to be avoided. Table 1 lists some of the typical functions that occur in a distribution operation. Many of these functions have unique subfunctions that must be balanced. Furthermore, many distribution operations have special unique functions such as count and weigh, special labeling and packaging, value-adding functions, returns and empty tote distribution. Distribution operations must have an even process flow with liquid-like properties to operate effectively and efficiently. In fact, a distribution center can be viewed as a series of connecting pipes, with materials and information flowing through the pipes much like liquids. As with liquids, if any one process operates at a slower rate, bottlenecks will occur, and workarounds creep into the process.
Workarounds come in many forms and may even appear like standard business practice if they continually exist. So how do you know when a system workaround exists? System workarounds typically require people to unexpectedly deviate from their standard routines. Routines are designed to keep people focused and achieve maximum productivity. When routines are broken, the focus is lost, and productivity is hindered. Workarounds create stress on the system including people. They can create chaos and the typical problems that surround chaos. Following is an example of a company that has a distribution system with existing workarounds:
Company XYZ distributes a product in a highly competitive industry that requires excellent customer service (orders shipped same day with 100% accuracy). Though the day-to-day business of Company XYZ has some peaks and valleys, it has been able to develop a system that meets the order fulfillment demands of its customers. On many occasions though, Company XYZ deviates from its efficient system to meet very high peak levels of business. Some of these spikes are predictable; however, some are completely unexpected. When these spikes occur, management and workers experience some level of panic, which results in a feeling of scrambling. There is no formal or documented process for these times, and decisions are made on-the-fly. Overtime is scheduled at the last minute as management realizes orders are not going to be fulfilled in the standard day. People are asked to deviate from their standard routines in an effort to do whatever it takes to get orders out the door. Materials sit idle at certain process areas, not able to keep up with the rest. Order accuracy and on-time delivery is lower during these periods, thus resulting in aggravated customers and increased returns.
The above example is generic but depicts a very common scenario. The following are some more specific examples of actions companies take when there are system workarounds:
Manually handling materials normally transported by automation For example, a specific process area has a system bottleneck, which causes products to back up on a conveyor system and hinders material flow. Staff in upstream processes needs to get material out of its process areas and reacts by manually transporting completed materials to downstream processes. The conveyor system has been bypassed; however, the materials manually transported are not shipped any earlier.
Discontinuing certain functions until the spike is over For example, discontinuing the replenishment of forward storage locations during peak times of the day in an effort to clear the floor for order pickers and maximize order-picking rates. Though it may make sense to stagger functional process shifts, shutting down a process during peak times in reaction to an unbalanced system will usually result in other problems such as lower fill rates.
Shifting people away from primary functions and/or routines to perform alternative temporary functions in an effort to meet peak demand For example, an unexpected large order is received, requiring 100% quality check and special labeling. The special handling process area has not been designed to accommodate such surges and is without enough staff or stations to handle the surge. In this case, the reaction may be to pull staff members from other functions and place them in makeshift special handling stations in order to process the order. An unorganized reaction has occurred rather than staff following a developed plan to handle such large orders.
The key for distribution managers is to understand the difference between disorganized and costly workarounds and those temporary efforts that are necessary, well-organized and have become a part of standardized operations to meet temporary business peaks.
The Causes of Bottlenecks
Change is one of the greater contributors to an unbalanced system and subsequent bottlenecks. Change in volumes, order mix, SKUs, operating methodology, customers, suppliers and other areas all have a pronounced effect on any operation. A good distribution company always looks at both the macro and micro effects that changes have on work balancing and material flow. Change in business typically requires some change to the distribution system, and the system must remain flexible to accommodate this change.
Systems design plays a significant role in whether a distribution operation is balanced or unbalanced. Though there are very sophisticated methods for measuring work, a basic approach is to determine the number of units of work processed per unit of time for example, cases per minute (cpm). All processes and functions should have approximately the same rate in order to achieve continual flow. A simple listing of functional areas can often quickly identify sources of potential bottlenecks in material flow and areas for improvement. For instance:
Process Rate (cpm)
Receiving 20
Replenishment 21
Order picking 22
Packing 15
Shipping 25
In the above example, it is clear that there is going to be a system bottleneck in packing. Regardless of the fact that receiving, replenishment, order picking and shipping can achieve 20 or more cpm, only 15 cpm are going to be shipped out of this distribution center unless costly workarounds are employed to address the system bottleneck, or the problem is addressed correctly and resolved.
It is important to understand that most distribution operations cannot justify the cost of systems designed to peak throughput levels, whether their processes involve only people or various levels of automation. Operations with a lesser variation between peak and average may be able to cost justify systems with the capability to handle rates closer to peak, while those with significant variations typically can only justify systems that accommodate rates closer to average. All systems, however, must be designed with flexibility and a plan to handle peaks as well as accommodate business growth. Without flexible systems and a plan for growth, companies with high seasonality will experience significant pain in the form of an unbalanced system, system bottlenecks, the inability to meet customer demands and service levels and escalating distribution costs.
Balancing Your System
There are a number of ways companies can balance their distribution systems and, more importantly, keep them balanced:
1) Know your processes inside and out, and document all relevant information including work description, rates and plans to accommodate peaks. Develop a distribution system process flow chart with rate capacities. Identify and address unbalanced areas.
2) Investigate how to isolate repetitive tasks. Sometimes there are savings in isolating repetitive tasks rather than combining tasks in a single functional area. For example, it is not uncommon for dunnage insertion and case sealing to be isolated for the sake of greater overall productivity.
3) Consider automation and technology. Some repetitive tasks can be automated and have a significant return through lower labor costs and greater throughput capacity. For example, automating the function of printing and applying labels can provide good return for the investment in the right application. Information systems can also be improved to benefit operations. Some examples are: scanning to replace manual data entry; utilizing advance shipping notices to benefit the receiving process and overall inventory planning; and using profiling and system modeling software to continuously ensure best warehouse practices.
4) Dont over automate. Automation can actually slow a system down if not used, applied and/or maintained properly. Automation can also be very limiting if misapplied. For example, certain types of automation, such as carousels and automated storage/retrieval systems, can only accommodate a certain number of personnel interfacing with the equipment. In comparison, shelving tends to have less of a limit regarding the number of persons who can access the media at any one time. In the right applications, carousels and automation are the best solutions; however, flexibility in staffing levels must be considered. When significant amounts of automation are utilized, ensure that they are continuously maintained because poorly maintained systems will cause severe systems problems.
5) Staff your process areas to accommodate required rates, and allow for scalability to meet peaks. Develop and document a plan for peaks, surges and growth.
6) Know your staffing options. For example, cross-training personnel can provide for flexibility and ancillary knowledge benefits (upstream processes will learn the effects of their actions on downstream functions). Know your temporary staffing resources ahead of time.
7) Understand the effect that your shifts have on productivity. Ensure that start times, breaks, lunches and shift overlaps do not have a negative effect on system balance and do not create system backups.
8) Continuously review alternative methods for each of the functional areas within your distribution operations as well as more comprehensive system methodologies. Test ideas that appear to have merit. For example, if you discretely pick orders, then review the potential benefits of batch picking. If you pick to totes, then consider picking to shipping cartons. If you check every order for accuracy, then try scaling back to 75%, and see if there is any difference in order accuracy. If a process is not broken, then break it, and try to make it better.
9) Ensure that waves are built properly and zones are balanced. If waves are built incorrectly, the proverbial five pounds of orders in a two-pound zone occurs. This can create a bottleneck that not only affects total order-picking rates, but it also negatively affects every downstream process. WMS packages today have wave-building modules designed to balance the zones within a system. If your WMS does not provide this, there are other distribution software packages designed to provide this function and/or enhance the manual wave-building process.
10) Continually pay attention to inventory profiling. Poorly slotted products result in lower productivity and throughput rates, which can result in system bottlenecks. Since businesses change, what works today may not work tomorrow.
11) Implement standards and a measurement process. Until you know where you are, you cannot determine where you are going. Know which measures best apply to your business, and make sure your entire team understands their importance. Though measures will vary by distribution operation, most of them typically pertain to productivity (orders processed per labor unit), costs (costs per line), inventory (fill-rates) and customer service (on-time delivery). Software is available to capture data and assist in the measurement of standards and the metrics evaluation process.
12) Evaluate the benefits of an incentive plan. Many companies have experienced significant increases in processing rates due to implementing incentive programs for their distribution staff. This can be an inexpensive way to eliminate a systems bottleneck. Ensure the plan promotes a healthy culture rather than an overly competitive environment and that it does not sacrifice quality for speed.
The benefits of having a balanced distribution system and stream-lined material flow include lower costs, optimized productivity, better accuracy and improved customer service. Removing the bottlenecks of your system will eliminate workarounds and increase profits.
Dan Graville is vice president of Western Sales at Fortna, Inc. For more information, please call 480-968-2429 or e-mail him at dangraville@fortna.com.
