Warehouse Activity Profiling
Identifying Improvement Opportunities in Your Warehouse Operations
By Catherine L. Cooper & Steve Mulaik
WHAT ARE ACTIVITY PROFILES?
Often in a murder mystery, detectives profile a killer based on the series of clues and facts. This profile gives them an idea as to the character they are looking for and enables them to predict his actions. Not that your warehouse is a suspect in any crime, but profiling its activities can be a great help in understanding its operations. A comprehensive profile based on historical and current data reveals characteristics that allow you to make decisions on storage and handling alternatives, slotting options, pick line requirements, system requirements, pick methods, and order releasing strategies. With more detailed analysis you can even predict how your warehouse will operate, providing you the invaluable opportunity of being able to prepare for the future.
Two main categories of profiles make up a basic warehouse profiling set: customer order profiles and item activity profiles. Customer order profiles represent the outbound activity, i.e., how the customers are ordering the products. Item activity profiles provide insight into viable storing and slotting options for each item within the warehouse.
Customer Order Profiles
The three most basic customer Customer order Order profiles Profiles are defined below.
- Order Mix Distribution: These distributions answer warehouse zoning questions such as “Should my fast, medium, and slow movers be zoned separately in the warehouse?” To answer this question, a distribution of the orders for fast, medium, or slow movers is compared with any combination of the three. If warehouse data indicates that most orders call for a mix of fast and slow movers, zoning the items by velocity will have order consolidation impacts that need to be considered. Order mix profiles are also used to analyze the percentage of order lines for full cartons, broken cartons, or a combination of the two. Analysis of this information provides options on storing full and broken cartons together. The goal is to determine what percentage of your customers is ordering full and broken carton quantities of the same item, before investing time and money in changing the current storage strategy.
- Order Increment Distribution: The order line distribution of the percent of a full carton ordered is beneficial when evaluating if the current packaging is in logical increments for the customer. For example, if results indicate that 90% of the customers are ordering ¸ carton quantities, the warehouse can consider effective alternatives. Options could involve first the supplier by changing the carton size, second the warehouse operations by breaking down cartons at receiving in order to save time during picking, and third, the marketing department by encouraging customers to order in full carton quantities.
- Order Lines Distribution: This distribution of the number of lines contained on each order is important when evaluating operating strategies. A warehouse with mainly one and two line orders would most likely have a very different picking strategy than a warehouse with many large multi-line orders. If your graph looks like figure 1 and you have a significant number of single-line piece-pick orders you may want to consider batch label picking these orders.
Item Activity Profiles
Item Activity Profiles are beneficial when analyzing products activities for the purposes of determining storage mode, product slotting, and facility layout options.
There are several types of item profiles, the three most basic are defined next.
Popularity profile- is a ranking of the items based on how often they are ordered or picked (frequency). Volume profile is a ranking of the items based on how much is ordered (cube movement). Finally, the item order completion profile displays the items ranked from most to least popular against the order set. This profile reveals the percentage of the orders that will be completed by a subset of the items and is valuable when conducting cost benefit analysis for improved productivity. The chart in figure 2 shows that 30% of the items complete 90% of all orders. This information is useful because it allows operations to make improvements, such as automation of a smaller area, yet still benefit 90% of all orders.
The best way to understand how effective profiles can be is by reviewing some examples. A tactical then a strategic profiling approach is discussed next.
A TACTICAL APPROACH: Slotting a Forward Pick Line
Three common methods for organizing and operating a forward pick area are: a straight pick line, a branch and pick zone, a serpentine pick zone. Each of these methods is shown graphically in figure 3.
In a straight forward pick line/zone, the most popular items need to be at the start of the pick path. This reduces the likelihood of walking to the end of the pick line to complete an order. In a Branch and Pick Forward Pick Zone, the more popular items need to be placed near the center aisle minimizing the time spent walking to the back of each cross aisle. In a Serpentine Forward Pick zone, the more popular items are placed in the aisles at the start of the tour, thus reducing the chances that an operator will have to visit every aisle in the zone to pick an order.
It is quite simple (if there are no seasonal or promotional items in the group) to take item activity information from a warehouse management system or inventory control system and sort it in order of popularity. Using this information to rearrange SKUs on shelves or in flow rack, can significantly cut an operators walking distance and therefore, reduce cost. This is “Activity Profiling” in its most simplest form.
If there are a number of promotional or seasonal items in your forward pick area, or you have introduced a number of new products, the analysis is a little more difficult. These items need to be isolated and decided on as to whether to include them in the “golden zone” within the forward pick area or not. This is done by generating a graph that looks like figure 4.
By graphing the % By graphing the percent of total days that each item is picked, it is easy to isolate the popular, seasonal/promotional items. Seasonal items tend to stick out like a sore thumb on these graphs. They will have high activity but the percent of days picked will be low. Seasonal/promotional items should be positioned towards the end of the line or not in the line when out of season.
A STRATEGIC APPROACH: Choosing a Picking Method for a Set of Orders
While Activity Profiling helps with many tactical improvements to your operations such as reslotting your forward pick area, even greater benefits can be achieved by using it as a tool to evaluate major changes to your operations.
Deciding on a method to release and pick your orders is an example of this type of analysis. Traditionally, most firms have picked orders using paper pick lists; however, many of the new warehouse management system applications allow users to pick orders in a variety of different ways such as batch, single-order, or multi-order. Furthermore, these systems support using labels, paper, or RF terminals to communicate such picklists. Activity Profiling can help make the decision on how to pick within a zone much clearer and easier.
For example, it is helpful for a facility using a warehouse management system to develop a chart like the one in figure 5. Managers can then document how they would like certain types of orders processed in the facility. Using activity profiling results can make completing this chart easier. For example, you should graph lines per order as shown in figure 1. In most businesses, in order to get as many orders out the door as quickly as possible, all single line orders are selected, grouped together and picked by label.
While people in case pick facilities are usually aware of this opportunity, mangers of pick and pack operations often overlook other options such as batch label picking orders. Label picking is fast; pick the item, place it in a shipping container, apply the label, and put it on the conveyor or pallet going to shipping. This process bypasses the packing operation altogether.
If you have many multi-line orders, profiling the Order Cube information is helpful. The results of such an analysis are shown in figure 6. Notice that in this scenario, almost 60% of the orders require less than a shipping cartons worth of material. If a large proportion of your orders are multi-line, consider multi-order picking those orders using carts with multiple slots, each slot dedicated to one order at a time. The same material handling concept can be applied to larger orders that are less than a shipping pallets worth of material. In this case a pull-train of pallets, each order having its own pallet, may be useful. Each of these approaches, reduces the travel time spent picking orders in the facility.
Now the big question is “How do I get started?” There are three main sources of data needed. Since much of this information is usually stored on computers, they are commonly referred to by their “MIS” name, i.e. item master records, order header and detail information, and inventory records.
What Data Do I Need?
The item master is a static file with each item’s stock keeping number, classification, weight, cube, pallet build, and dollar value. The order header information contains the name of the customer receiving the order, the customer’s location, and other information that is common to all lines on an order. The order detail record has information specific to a single line on an order. It contains the item ordered, the date ordered, quantity, etc. Inventory records provide the quantity on hand of a given item, in a given location, at a specific time.
Most of this data is available from your warehouse management system or inventory control application. Unfortunately, it is often inaccessible by PC users. In these situations, the IT staff should transfer this data from the mainframe to the PC via a zip disk, CD, or the web.
It is important to collect at least 6 months and preferably a year’s worth of data so that the profiles account for seasonal items. However, if there is little seasonality in your business and the products you ship are relatively static, less data may be used.
How Do I Manipulate the Data?
In most cases the data will require some manipulation and cleaning prior to analysis. Obsolete items will need to be removed and incorrect item numbers corrected before the data set is loaded. The clean data is loaded into a PC database product such as Microsoft Access. Once the tables are created in the database package, they are exported to spreadsheet software such as Microsoft Excel for sorting and ranking of the data to create the profiles. Typically, spreadsheet tools are better than database tools for sophisticated analysis such as developing a histogram of Order Cube. Figure 7 outlines the data flow profile process.
Activity Profiling Schedule Estimates
A profiling project’s schedule depends on data availability, fidelity, and the amount of analysis performed. While most projects can be completed in four to five weeks, the “X Factor” is the time it takes to identify where the data resides and exporting it to a tape or zip disk. Once this is done, the database loading process takes typically one week. This time includes creating the tables, running some preliminary queries, then exporting results to a spreadsheet for graphing or further analysis.
Next, the preliminary results need to be reviewed by your improvement team. This stage can take a couple of weeks. The first step is to review findings related to previously identified opportunities followed by selecting and evaluating new improvement areas. Overall, the profiling process should take about three to four weeks to determine reasonably good opportunities.
Profiling is a Great Tool If·
Benefits derived from a profiling project can be significant. The payback can also come quickly; however, watch out for three pitfalls. First, many companies find that ideas that were being considered before the analysis are no longer viable based on analysis of the data. More often than not, the real opportunities tend to be initiatives that were not thought of prior to the start of the project. The lesson here is to keep your eyes open and your mind prepared for doing something different.
Second, a word of caution about spending too much time on the analysis and getting mired in the data. A picture is worth a thousand words but too many pictures and numbers may prevent you from seeing the forest because of the trees. It is important to have a clear understanding of your major warehouse goals and to decide on the data diagnostic set that will provide insight into your objectives.
Finally, activity profiling is not to be considered a one time event. As your customers, products, and operations change, re-employ this tool to confirm current practices and look for new improvement opportunities. In this way, profiling can earn you a return for years to come and your warehouse can continue to provide you with clues for operations improvement opportunities.
The article above first appeared in Operations & Fulfillment magazine.
Source: The Progress Group