If different tasks proceed at different rates, try to arrange work steps so that the faster process is executed before the slower one
If your WMS is building carts, investigate basing the process on cubic capacity rather than on a set number of orders
An ounce of prevention is definitely worth a pound of cure this year, if, like many businesses, you are striving to reduce operating costs. In a distribution environment, wringing maximum efficiency from operations with the least capital expenditure means reducing labor expense.
Broken-case picking typically accounts for the greatest head count in a distribution center, and it is in this area that you can most effectively assume the role of the “picking doctor,” check for danger signs, and prescribe preventive action. Take out your virtual operations stethoscope and run this quick diagnostic test:
- Do you conduct operations according to written procedural guidelines?
- Does each employee follow the same work steps in the same order, or does each person have his or her own way of doing things?
- How often do you do “work-arounds” or “fool the system” to accomplish everyday tasks?
- Ask your supervisors and employees why they perform certain jobs. Do they have clear explanations, or do they reply, “That’s the way we’ve always done it”?
If the answers to these questions are unclear or unexpected, then you have probably already recognized the opportunity for improved efficiency in your operation.
“Tolerance stacking” is a term used in manufacturing to describe the phenomenon of small variations in individual component parts production adding up to a point where the end product does not perform to specification. This phenomenon occurs in most operations over time — small variations, which seem insignificant when considered separately, creep into the process. Taken together, such deviations create a significant drag on efficiency. In a fulfillment operation, it is the picking doctor’s task to ferret out these latent inefficiencies.
In any operation, responsible managers have tried to keep pace with the changing business environment — instituting a work-around here, an offline process there, a one-time project that has become a part of daily business, new demands that led to a large influx of new hires or temp labor.
The doctor is in
To expose some of these inefficiencies, just walk through your operation as though you were seeing it for the first time. Resist the tendency to skim over familiar areas without questioning your assumptions. Some symptoms may not be obvious without closer examination.
To perform this review, let’s resuscitate the doctor/patient analogy. Doctors diagnose patients’ illnesses through observations, specific questions, and tests. After the diagnosis, he/she will prescribe a course of treatment. Here the “picking doctor” will diagnose common productivity problems in picking operations and offer prescriptions for boosting efficiency.
PROFILING Diagnosis: Spend some time observing the pick aisles and talking to employees.
- Are certain pick aisles congested?
- Do the pick paths seem too long?
- Do some bins need to be replenished every wave instead of every week?
- Are some bins almost empty and others overflowing?
Rx: If the answers to these questions are “yes,” then it is probably time to examine your profiling and slotting effort. (See “Daily Grind” on page 41.)
EMPLOYEE PERFORMANCE Diagnosis: Observe employees’ job performance across the operation.
- Are some employees more productive than others?
- Do the more productive employees simply work faster?
- Are they better organized, or do they use different procedures?
- Do some employees wait for work, while others work continuously?
- What is the error rate for data entry?
Rx: If improved productivity is simply a matter of faster work rates, then incentive and employee recognition programs are effective motivators. Implementing a pay-for-performance program is a major project, but the cost savings generated are well worth the effort.
Use employee performance variances to your advantage by incorporating the best practices into your standard procedures. Involve employees and supervisors in the best-practice brainstorming process. Allow time for retraining in new methods and procedures.
If different tasks proceed at different rates, causing some employees to stand idle while others work continuously, try to arrange work steps so that the faster process is executed before the slower one. This will create a work-in-process buffer between these activities, trading staging space for wait time. Alternatively, supervisors and leads can shift employees between jobs to minimize idle time, or jobs can be combined in an effort to equalize the work rates.
For work steps that involve keying information into your warehouse management system (WMS), a bar code scan in place of the keypunch operation will reduce time and errors.
If pickers are spending time sorting and grouping documents, consider shifting this job to a clerk so it will not hinder the utilization of your direct labor. Better yet, software modifications could automate the sorting and grouping of documents as they are printed, eliminating this step entirely.
WORKFLOW AND LAYOUT Diagnosis: Note operations requiring an employee to walk from one point to another.
- How far must employees walk to deliver pick assignments to packing?
- How do employees receive their picking instructions?
- How are empty totes (or pick carts) obtained?
- How are supplies obtained? Replenished?
- How is trash disposed of?
Rx: There is no magic elixir for curing workflow and layout problems. Installation of conveyors, different types of storage, reconfiguration of existing pick aisles, system modifications, and changes to the process itself are all options to be considered. Some of the most effective changes can be the simplest. For example, consider the following scenario in a tote-picking operation:
If employees must walk to a central staging area or to the end of an aisle to retrieve empty totes, have re-stock workers place empty totes in several locations throughout the picking area.
It may not seem like much initially, but think of the time it takes a picker to walk twice the length of a pick aisle (round trip) to retrieve a tote stocked at one end of an aisle. Multiply that time by the number of totes each picker handles every shift. If the maximum round trip time is 90 seconds, and employees pick 20 totes per hour, four totes at a time during an eight-hour shift, each employee makes 40 trips to pick up totes, spending 60 minutes of each shift retrieving totes.
It is more likely that the average travel time is significantly less than 90 seconds. However, a 45-second average travel time still amounts to 30 minutes of unproductive time per shift for each picker. At this rate, cumulative annual non-productive time for 50 pickers is 6,250 payroll hours, or 125,000 totes worth of picking wasted in walking. Placing empty totes at each end of the aisle reduces that time by half. Situating an additional stack of totes in the middle of the aisle reduces that unproductive time by 75%.
CART PICKING Diagnosis: Analyze your picking process.
- How are orders grouped into batches?
- Who sets up pick carts?
- Are aisles significantly congested?
Rx: Software modifications may be in order in your cart pick area if orders are manually grouped onto carts rather than built by your WMS based on pick path. If your WMS is building carts, investigate basing the process on cubic capacity rather than on a set number of orders. This will maximize the work content on each cart.
If each picker builds cartons and sets up his/her pick cart after completing the previous cart, consider creating a separate job for cart setup. Let’s say a picker takes an average of 20 minutes to pick one cart and two minutes to set up a new cart. During an eight-hour shift, the average picker should be able to pick just under 22 carts’ worth of orders.
Now eliminate that two minutes of set-up time from the picker’s job by hiring unskilled temp labor at $7 per hour to perform the task. The picker can now pick an additional two carts per shift. But you have spent an additional $56 on the temp, so where’s the savings? The temp can set up 30 carts per hour, so he/she can set up carts for 10 pickers. That translates into an additional 20 carts per shift, which is almost one full-time employee’s worth of picking. If the average loaded labor cost for a skilled picker is $10 per hour, you are effectively getting the output of an additional skilled picker at a discount of $3 per hour. If five temps support 50 pickers, that is an annual savings of $30,000 over maintaining 55 full-time pickers.
If aisle congestion is a significant problem, establishing “hand-off” zones to pass carts from one picker to another could help alleviate the situation. This change will also require procedural changes in picker accountability, as a single cart will be passed to multiple pickers. If aisle congestion is the major problem in your operation, then it may be time to consider a true sequential zone pick operation.
PICK AND PASS Diagnosis: Examine how each picker performs his job.
- Does each picker examine every tote, sometimes picking, sometimes not?
- Is each picker assigned to a fixed zone, sometimes overloaded, sometimes idle?
Rx: Pick-and-pass operations depend heavily on layout and systems for efficient operation. Improving efficiency in a poorly designed operation will require a significant investment in both material handling equipment and systems.
If your layout routes every container through every zone, then installing some conveyor controls and equipment to ensure that containers are routed only to zones requiring picks will dramatically improve the efficiency of your operation.
Fixed pick zones will result in variable amounts of work being assigned to the zone. Software modifications or implementation of a picking system such as pick-to-light to support variable pick zones based on pick density will even out the work content among pickers.
As you are brainstorming innovative procedures, work methods, and rack changes, and developing your wish list of new software to improve efficiency, pause periodically and make sure your picking flow still makes sense. Do not optimize one process at the expense of another. Also, consider how new processes will affect operations up- and downstream of picking. Will you require faster replenishments? Will vastly increased throughput overload the packing operation? Just as your doctor considers your total health when prescribing treatment, you should consider the impacts on the overall facility when making picking operation changes.
Although modifications to existing operations may be perfectly logical and should be implemented on their own merits, never lose sight of the bottom line. Even strictly procedural changes have costs associated with implementation (training, design, delivery, ramp up to new productivity).
Gaining approval for new equipment capital expenditures can be difficult in the best of times. In today’s fiscal climate, conducting a thorough cost-benefit analysis is imperative. Determining the potential savings generated by process improvements can be tricky, but as the examples have demonstrated, by starting at the rudimentary level and building from there, you can arrive at fairly accurate annual savings estimates — and the results may surprise you.
When implementing process improvements on the floor, the strongest case for change will clearly communicate the appropriate goals and illustrate the path to achieving them. Some employees will jump on the bandwagon immediately; others may take longer or even refuse the change altogether. Recognize that different people react differently to change. The way your organization handles those different reactions can have a dramatic effect on how smoothly the implementation proceeds and how quickly gains are realized. Help your employees accept change by explaining the plan, making them part of a team so they do not feel isolated, giving them the right tools and training, and celebrating their successes.
Chris Merritt is a member of the direct-to-consumer practice of Atlanta-based Kurt Salmon Associates. He focuses on fulfillment centers, contact centers, and the systems that drive them for direct marketing and retail. Merritt may be contacted by phone at (404) 898-7916 and by e-mail at email@example.com.
Marc Bessho is a member of Kurt Salmon Associates’ fulfillment technology practice. He specializes in the integration and deployment of warehouse management systems (WMS) and associated material handling control systems for retail, catalog, and Internet distribution projects. Bessho may be contacted at firstname.lastname@example.org
No one would argue that the location and amount of product in the active area is a key to efficient picking. All too often, however, the assignment of product to pick slots is based on which slots are empty. Over time, this practice diminishes efficiency, as limited, high-volume pick slots are consumed by slow-moving inventory or fast movers are slotted into low-capacity storage requiring excessive replenishment. Continuous profiling of product demand characteristics and appropriate slotting will increase space utilization and help ensure and maintain efficient picking.
The first step in profiling is to ascertain the highest-velocity SKUs. Divide picking data into ABC categories for fast, medium, and slow movers. Just obtaining the category data will facilitate some decisions about storage type. You might assign fast movers to pallet locations, medium movers to case flow slots, and slow movers to static shelving. The problem with ABC profiling is that it does not account for space requirements. Cubic velocity takes these space requirements into account.
To calculate cubic velocity for a SKU, multiply the velocity (units per day) by the unit cube to determine the cubic velocity in cubic inches per day. Now, multiply this number by your desired stock level (how many days’ supply in active reserve). This final number is the cubic inches of storage space required in active to maintain your desired inventory level without excessive replenishment. Demand forecasts from merchandising will form the basis for slotting decisions on new SKUs with no historical picking data. To account for inaccurate forecasts and demand life cycles and to maintain optimal slotting, profiling must be a continuous activity.
The immediate effect of storing product according to its cubic velocity profile is that replenishment is based on its demand profile and not on the capacity of its slot. This will improve space utilization and, depending on your WMS and picking strategy, will lead to increased pick density for a pick zone and/or reduced repeat visits to pick bins across a shift.
Given appropriate data and analytical tools, SKU profiling can also take order characteristics into account to determine if certain items are ordered together or in various combinations. Such items can then be slotted in proximity to each other or, if warranted and supported by your WMS, certain SKUs can be slotted in multiple pick zones around different “SKU clusters” to minimize the pick path.
Developing hard-cost savings data on staffing and implementing a rigorous profiling/slotting program may be difficult; however, the effect of a poorly managed effort will be obvious in picking and replenishment. Stand-alone and integrated software programs are available to help with profiling and slotting, but initial efforts can be based on spreadsheet models fed with data from your WMS.