When Murray Feiss Industries Inc. opened its 70,000 square-foot distribution center in the Bronx, the manufacturer of residential lighting products faced a severe space crunch of its own making. The family-owned company wanted to remain in New York City, but municipal zoning regulations set tight limits as to how big the facility could be.
To achieve its goals of maximizing distribution in a rather minimalist amount of space, Murray Feiss relied both on old-fashioned street smarts and next-generation artificial intelligence to develop its DC.
The facility as designed is very narrow, with multiple pick areas to accommodate varying demand. Items are classified based on a combination of size and selling volume. “Up to a certain cube, we put the A items — the best sellers — in a six-level pick tower,” explains Rita Hoffman, vice president, operations with Murray Feiss. Conveyable larger best sellers — the B items — also are in a convenient pick area. Less frequent selling C items are in a less accessible area. And there are separate pick areas for items too large to be conveyable and for broken-case picks.
“When we receive goods, we enter data in our warehouse management system (WMS),” Hoffman says. “The rules-based system classifies items and knows where to locate them,” notes Hoffman.
The WMS, supplied by Manhattan Associates Inc., includes an artificial intelligence-based component that optimizes inventory placement and warehouse space utilization.
“We plan picking through the WMS in waves,” Hoffman continues. “The WMS prepares the labels in logical sequence for efficient picking and we batch pick all orders in a wave. Then the conveyor system sorts by order, per WMS pick instructions, pre-configuring the pallet for each order. The system’s planning module picks up all the instructions.”
Murray Feiss prefers to do all its replenishment based on thresholds which, Hoffman points out, should never drop to zero. “The WMS might determine, for example, that we need a minimum of two pallets of item X in area Y and the night shift does the replenishment. As we do the actual order planning, we set up demand replenishment when needed. We do that as seldom as possible.”
With four new-product introductions scheduled per year, Hoffman’s team reviews pick tower slotting with each introduction. “In addition, if an out-of-stock item comes in — meaning lots of picks for a short time — we’re flexible enough to accommodate it in the pick tower,” she adds.
The new system is designed to handle up to 10,000 boxes per day — the level Murray Feiss expects to achieve within its five-year plan — and it easily manages the current 6,000 boxes per day.
It’s critical to re-slot on a regular basis to accommodate seasonal changes, aging products, or new products, says Geoff Sisko. As senior vice president with Gross & Associates, a consulting firm specializing in material handling logistics, Sisko notes most people don’t re-slot frequently enough. “You have to review line items and develop criteria. Even though software can do it faster than a person, it still requires human intervention. You need to ask why an item is being re-slotted. Was it out of stock or marked down?”
“The baseline of slotting is understanding the business dynamic, not just the lines,” concurs John Giangrande, senior account executive with Fortna Inc., a design/implementation firm specializing in distribution center (DC) design. “We use a DC modeler to look at a year’s worth of data. We measure by day or week to set the table for slotting redesign or a new DC. When slotting, look at movement and cubic volume, then cross reference to orders. Know what SKUs help complete the largest number of orders.”
“With effective slotting, you can increase space utilization, reduce replenishment labor, reduce picker stockouts and increase picker throughput,” claims Norman Saenz, manager of logistics with Carter & Burgess Inc., a consulting firm specializing in distribution, retail delivery and logistics. “Look at volume per day of an item and put it in a location that can handle three to five days’ volume. Identify fast movers and put them in a forward pick area instead of general storage. This approach cuts travel time and vehicle traffic,” he adds.
However, Saenz notes, you have to balance the physical pick area with replenishment labor and picking labor. With a smaller pick area, pickers spend less travel time to locations but there are fewer days of inventory on hand. Generally, space is cheaper than labor, so Saenz prefers a smaller storage area and a large forward pick area.
Regardless of how you use the space, if you don’t have a plan for product slotting in the pick area, simply throwing technology at productivity is apt to be disappointing, suggests Saenz. A large retail customer he rescued had installed pick-to-light for every SKU, even slow-moving C and D items. With no plan to slot properly, C items — which only move once a month — were in the forward pick area, which meant they were getting in the way of the fast-moving A and B items. The efficiency of the warehouse staff, consequently, was being impeded by the slower-moving items.
A common mistake is implementing islands of technology that don’t talk to each other, says Giangrande. Lack of communication between zones or between carousels and conveyor system, for example, could mean increased stockouts and manual data entry.
Improper storage media is another potential roadblock to efficiency in the pick area. “Do the storage media handle goods that come in on pallets and go out in cases or eaches?” asks Giangrande. “Maybe you need a pallet flow rack of a single SKU for large volume items. Can you group SKUs in a small lineal foot traffic area so a picker can make lots of picks with minimal walking? Can you stock and pick concurrently rather than competing for space and resources? If you fail to allocate enough space per SKU, it could mean a lot of replenishment activity or frequent stock outs,” continues Giangrande.
In slotting product, it is critical to think through the entire process and include the human element. Normally, companies slot frequently picked product together, states Sisko, but there are always exceptions. One company he’s worked with — a distributor of fishing lures — arranges product by vendor in its facility. Because its catalog is arranged by vendor, orders also tend to be grouped by vendor.
Even where slotting rules say put the fast movers together in the golden zone — at a height between a person’s shoulder and hip — if you group all the fast movers together, multiple pickers will get in each other’s way, notes Sisko. Also look at weight. Crushability is product specific. And flammable product must be packed separately, so slot them separately.
Whether variables are many or few, “execution is paramount,” says Bill Horrocks, director of logistics engineering with third-party logistics provider (3PL) USCO Logistics. “Yet people get caught up in the glitz of new technology.”
There is a constant flow of new products designed to enhance the order-picking process, notes Marc Mitchell, transportation practice director with systems integrator Enterprise Information Solutions. The most common of these are pick-to-light, radio frequency and voice-activated devices.
Using a hand-held or wrist device, radio frequency (RF) transmits to and from a base station. The operator scans the product’s or location’s bar code, and the device indicates which item to pick. The operator then scans the item to verify the pick, yielding a high degree of accuracy.
Voice-activated picking directs the picker to the location, tells the number of each item to pick, and verifies the pick by reading the location number and repeating the count. In case pick, it’s more effective than scanning because both hands are free and it both tells and verifies the location. It also describes the item for comparison with what’s there, explains Sisko. Voice can be set up in any language as long as it is programmed to recognize the responses.
Where pickers travel longer distances, voice is competitive. In fact, Saenz notes, picking cases off pallet flow racks to conveyors using voice technology can yield a 15% to 20% productivity gain over paper.
“Pick-to-light technology can be a good choice for piece or case picking,” says Sisko. One variation is put-to-light, in which a retailer has a box for each store. If the picker scans a box of ties, for example, the put-to-light system tells the picker how many ties to place in each box.
With pick-to-light or RF, information is available immediately to the WMS. Ideally, the WMS ties in to other technology, as these methods can help streamline picking, make it paperless and keep information systems live in real-time.
In open case picking, pick-to-light yields a high level of productivity, says Saenz. Individual pickers can average 270 lines per hour. But voice is making headway in that environment. Suppliers are working toward a combination of scanning and voice, he adds.
All these technologies give improved speed and a higher degree of accuracy over paper systems. They also have the effect of logging work by time consumed, thus measuring the efficiency of the picker.
Investing in new technology may improve the efficiency of individual pickers, but the real savings when you replace a paper-based system come from eliminating pre- and post-picking activities. One company that worked with Carter & Burgess achieved its ROI within two years by eliminating three pre-pick positions and saving the cost of paper and time spent entering data. “With paper, someone has to collect the paper and input data manually,” Saenz points out.
Productivity with RF terminals may not be as high as other technology, but RF has the highest marks in accuracy, claims Saenz. If you use RF in replenishment to verify product is in the right place, pick-to-light or voice technologies come closer to the accuracy of RF picking.
Weigh current needs for speed or accuracy, but also ask, “Will the selected technology support more locations, SKUs or volume?” With RF or voice-activated picking, you simply add equipment or people to handle more SKUs, Saenz notes. In pick-to-light, if you add locations, you have to add or reconfigure the lights. However, to handle more volume, just add people to pick-to-light. With the other systems, you have to add equipment, Saenz says.
Because your business will continue to change, don’t expect slotting to be a one-time task. Plan for change and understand your business needs to achieve maximum efficiency in slotting and order picking. LT