Batch vs. Wave Picking

Jun 01, 2006 9:30 PM  By

In the distribution center, walking from one location to the next accounts for 60% of a picker’s time. Reducing walking time, then, is all but guaranteed to improve efficiency. That’s the rationale for batch picking and wave picking.

But how do you know which method will be the most efficient for your warehouse? Largely by analyzing your order profile (how many lines per order you typically sell) and your SKU commonality (how many types of SKUs you offer).

First things first: defining the terms

In batch picking, groups of orders are picked at the same time to minimize repeat visits to the same product bin location. An order management system (OMS) will arrange a group of orders in the most efficient picking sequence.

Alan McDonald, systems consultant for Mason, OH-based supply chain services provider Forte, uses a grocery store analogy to explain batch picking. Say you’re going to the store carrying two shopping lists, one for you and one for your mother. You could shop for all the items on your list first, working your way from one end of the store to the other, then shop for all the items on your mother’s list, repeating the same process. Or you could aggregate the two lists so that you make just one pass through the store. When you’re in the produce aisle, for instance, you’d pick up both the carrots that are on your list and the apples on your mother’s.

Now suppose you and two friends are doing the grocery shopping for all 60 households in your neighborhood. You could divide the lists into three batches of 20 so that each of you could make a pass through the store. Or you could aggregate all 60 lists, then divide the shopping into zones: One of you focuses on just the refrigerator/freezer products on one end of the store, while another handles just the produce at the other side of the store, and the third picks just the dry goods in the middle of the store. This would be wave picking: multiple pickers in zones throughout the warehouse picking items in those zones only for an entire “wave” of orders.

The benefits of batching

In small-order environments characteristic of many direct businesses, batch picking can boost your pick rates from 60-70 lines an hour for single-order picking to 200 lines an hour or more, depending upon the order profile and the average cubic size of your orders. For example, a contact-lens marketer, whose product is roughly the shape of a deck of playing cards, could fit perhaps 400 contact-lens cases on a pick cart. Conversely, if you’re picking bulky items such as duvets, you may only be able to fit a few orders on a cart.

In fact, because batch-picked items are typically transported to the pack station via cart instead of via conveyor as is more typical of wave picking, there’s a low cost of transport. Pick carts typically cost $300-$700, depending on their complexity.

For the most part, says Sam Flanders, president of Durham, NH-based Warehouse Management Consultants, any time you have a high concentration of SKUs over a larger area, you want to use a batch-picking strategy.

Framingham, MA-based operations consultant Wayne Teres says that many companies opt for batch picking if SKUs are in densely populated locations (for instance, a jewelry merchant with 5,000 SKUs in one compact area) and are easily hand-picked, or if the company does not have a conveyor or a tilt-tray sorter.

Companies whose orders tend to have similar numbers of lines typically create one-size batches; each batch would consist of, say, six orders. Based on the characteristics of the orders, you can also create multisize batches, based on aggregate number of lines or cubic volume. Say you define a batch as consisting of 20 items. One batch could consist of 10 orders, each of which has two lines, while the next batch could consist of just four orders, one of which has 10 items. By taking a varied approach to the number of orders per cart in your batch, Teres says, you’ll likely have less pick-cart traffic on the the distribution center floor, allowing for more throughput and increased productivity.

Making waves in the DC

For larger companies, however, batch picking doesn’t always reduce pick-cart traffic enough. When Freeport, ME-based outdoor gear, apparel, and home goods cataloger L.L. Bean was doing batch picking years ago, one of the biggest problems was that its workers were literally bumping into one another with their slotted pick carts during its peak season, says spokesperson Rich Donaldson.

“Batch picking works for smaller operations,” Donaldson says. “We just outgrew that method.” The company opted for wave picking when it opened its 660,000-sq.-ft. warehouse in 1996.

Wave picking is appropriate in companies with large numbers of SKUs. And because items are not sorted down into individual orders when they are being picked and therefore tend to be more uniform in terms of size and shape, the picking cart can typically hold more items than with batch picking, enabling more orders to be picked during a pick tour.

If you have a warehouse management system (WMS) you can create order waves to see how much labor you’ll need to allocate and how much time it will take to complete each wave. And as the wave works its way through the DC, you’ll be better able to allocate how many downstream operators you’ll need once the picking process is completed.

This bring us to a drawback to wave picking: Once the items for the wave of orders have been picked, the orders then have to be in effect collated downstream, sorted down to the individual order.

Like the batches in batch picking, the waves in picking can vary in size. You can also program the WMS to generate specific kinds of waves, such as sorting the orders by parcel carrier. Creating waves of orders that are all shipping via United Parcel Service, for instance, can ensure that those orders are ready when the drivers are scheduled for their pick-up. WMS providers such as Atlanta-based Manhattan Associates; Eden Prairie, MN-based HighJump Software; and Waukesha, WI-based RedPrairie feature good wave-reporting capability, says Forte’s McDonald.

Randy Scalise, vice president of fulfillment for Warren, PA-based apparel and home goods mailer Blair Corp., says that if a significant percentage of your business is multiple-line orders, wave picking allows you to service those orders more efficiently by taking advantage of the technology of an order/packing sorter. Most operations using wave picking have a tilt-tray sorter system to help transport the goods through the distribution center. At a cost of $1 million or more, though, tilt-tray sorters are feasible only for the largest operations. Less expensive options include a simple belt conveyor, which costs $180-$250 per linear foot, and a bomb-bay sorter, which costs about $40,000.

At L.L. Bean, once items are wave-picked from their stock locations, they are introduced to a sortation stream for order fulfillment. The picked items are placed on the conveyor system, where each item is scanned for barcodes. The scan is used to match the product with the location of the SKU’s appropriate packing station, where it is delivered by the tilt-tray sorter. The employee at the pack station, who has the list of items for each order, then verifies the SKU with information on the order.

Fixed waves, dynamic waves

Within the wave packing method there are two options. With fixed-wave picking, orders aren’t sent off to be packed until the entire wave’s worth of items have been picked. With dynamic-wave picking, each order is sent to the packer as it is completed.

The drawback to dynamic waves, according to Flanders of Warehouse Management Consultants, is that you need a dedicated employee to watch the sorting line for the completed orders. Once the order is completed, the employee must replace it with an empty carton for a new order: The sorter can’t deliver new orders to that location until the old orders have been cleared away.

And because you never know where the next completed order is going to pop up, the employee will be doing a lot of looking and walking up and down the sorting line. On larger systems, you’ll need multiple people watching for completed orders.

Sorters at a glance

Sorters bring the product to the order. In operations where there is a great deal of commonality, this means that the fast-moving SKUs can be brought to the sorter and inducted all at once; the sorter figures out which orders get the item and how many items to dispense. Even where there is little or no commonality, a sorter can enable the picker to select many orders at once to a single vehicle and then dump all the picked items into the sorter without regard to which order needs a particular item. Among the most commonly used sorters:

  • Shipping sorters can help get your products to their destination more rapidly. In a case operation, the sorter moves the cases to a shipping lane. In a parcel operation, the sorter will sort the product to one or more carrier lanes. Sorters range in speed from very slow to more than 100 cases per minute. In most parcel operations a sorter in the range of 15-30 cases per minute is adequate. You can usually purchase a shipping sorter from a conveyor company, although some operations have used tilt-tray sorters as shipping sorters.

  • With a tilt-tray sorter, the product is placed onto a tray (either automatically or by a person). When the tray reaches the order to which the product belongs, it tilts, dumping the product down a chute or into the order container. Tilt trays are among the fastest sorters available.

  • The bomb-bay sorter drops the product directly over the tote or the carton. It can be used for relatively small products as well as for larger but soft items such as apparel. The bomb-bay sorter tends to be simpler in construction and mechanical design than other types of sorters, so it is quite reliable and often the least expensive of the sortation options.

  • A cross-belt sorter is a conveyor sorting system that uses a series of carriers mounted on a conveyor to sort materials. Each carrier has a small belt conveyor mounted on top of it that runs perpendicular to the main conveyor. When it arrives at a sort point, the conveyor on the carrier will spin, moving the materials to the side of the main conveyor (usually onto another conveyor, down a chute, or into a container).

Source: Warehouse Management Consultants; www.inventoryops.com

Batch up your singles

In batch picking, groups of orders are picked at the same time to minimize repeat visits to the same bin location. Your order management system (OMS) arranges groups of orders in picking sequence, allowing the order picker to visit a picking location and select all the items at that location for the entire group of orders. Typically the picked items are then distributed to the appropriate orders.

When programming your OMS for batch picking, Framingham, MA-based operations consultant Wayne Teres suggests creating separate batches of single-line orders rather than mixing single-line orders with multiline orders. “It’s a cubic volume issue when filling up pick carts,” says Teres.

Because single-line batches contain only one item per order, once you’ve picked the item you’ve completed the order. You can place it in a pick cart and move on to the next order. With multiline orders, you generally run out of space with your batch carts much more quickly. Say you have a 1,000-order day; if the batch size was 10 orders per batch, with each batch including both single-line and multiline orders totaling 40 items, there would be 100 batches in the day. But if 20%, or 200, of those orders are single-line orders, you could store on one cart 40 single-line orders. There would be five batches of single-line orders and 80 batches of multiline orders for a total of 85 batches for the day, a reduction of 15 batches. And, Teres says, fewer batches translates to less time required to pick the orders, increasing productivity and reducing costs.
MDF