That inventory is money is not exactly a news flash. The sooner managers and employees understand that simple fact, however, the sooner inventory management will improve. Inventory, like money, has to keep moving for a supply chain to be cost effective. Miscounting money or inventory does not mean that it's not there—but it might as well not be.
Inventory management specialists have long lists of ways to keep inventory and data flowing through a manufacturing process or distribution center. Here are some of their more frequent observations, not necessarily in order of importance.
Get rid of the paper
The easiest way to create inventory management problems is to continue to use paper. Using paperwork to manage inventory is almost a guarantee to slow things. At the least it's a sure-fire way to create inaccuracy and confusion.
"What order pickers need is realtime data," says Tim Wills, director, Peak Technologies (Columbia, Md., www.peaktech.com). "Without automating data collection there will be piles of paper waiting to be keyed in, along with all of its negative repercussions."
Without real-time data, order pickers might select items previously allocated but not keyed in. And on the inbound side, orders can go unfulfilled because inventory sitting on the receiving dock has not been entered into the system.
"It's about visibility of the goods," says Wills. "Without active collaboration back to your trading partners your visibility forward [toward your customer] is limited. And that visibility is achieved through automated data collection."
Whether a company can automate its inventory management depends on its existing IT structure. And the level of automation chosen often depends on budgetary constraints.
Getting things in sync
Hampered by a legacy EDI system? Is data available only sporadically? What happens when trading partners begin asking for more frequent updates?
The critical piece to improving inventory accuracy, says Rick Nucci, chief technology officer, Boomi Software (Berwyn, Penn., www.boomi.com), "is to synchronize existing systems by integrating them with each other in real time."
Nucci says inventory accuracy spans three or four systems within a company, not just what's happening on the shop floor. Currently, the evolutionary step companies take is to tie back inventory data to all of their sources. "Maybe they have an EDI connection, now they want to go to the next level with real-time automation," he says, "and they don't want, or can't afford, to start with all new software."
A batch data collection system that gives data about what happened " today" is not the answer because it's not real time. By directly integrating with suppliers and customers systems, a company can improve information flows from procurement through order fulfillment. Nucci recommends starting inventory data management improvements with a discovery process. Managers should look at where they are now and what could be possible.
"At the end of the day," he says, "technically speaking, integration is about quickly and easily being able to connect systems and exchange data.'"
Kevin Hume, a consultant with ESYNC (Toledo, Ohio, www.esync. com) lumps inventory management problems into three general categories: poor processes, poor discipline and poor IT design.
Implementing new processes without consideration for the impact to data collection, and how resulting inventory inaccuracies might happen, is at the top of his list. For example, changing from collecting inventory information via the SKU, to collecting data via SKU and lot number. There might be good reasons to collect the lot number, but the lot number is not always consistent with the product in a box. The impact on inventory data management occurs when an operation follows strict FIFO (first in first out) procedures. If the FIFO window is defined by a period of time, and all lots are considered equal and commingled, a WMS may not be able to support that strategy.
Collecting excess data is also a problem, one solved by following lean manufacturing philosophies. "As more folks strive for lean principles," says Hume, "they find it easier to identify underlying poor processes that contribute to inventory inaccuracy." As an example he cites collecting product serial numbers on inbound product when all that is really required for the customer is to record the serial number on what is shipped. Eliminating excess data eliminates excess work.
Can RFID help or hinder some of this data collection challenge? "Actually RFID can be an enabler of collecting-excess data," says Hume with a laugh. "I view [RFID] like a 10-foot rope: You can swing from it or hang by it." The challenge is determining what managers need to get out of the data that RFID generates.
Poor discipline is typically the result of complacency, says Hume. It can have a major impact on inventory accuracy. "It's not an uncommon practice in warehouses for employees to stage pallets, then confirm putaway by keying information into an RF device. Later in the day they physically move those pallets to the location they had earlier indicated." For the person allocating or picking that product, it appears in the system that the inventory is in place, when actually it's still sitting on the dock.
"One work around we've developed," says Hume, "is to create an alias location label with check digits. That way the person cannot just key in the location. The pallet has to be placed in the spot and the label scanned before the product is released for picking."
Another area to watch, where operator discipline problems impact inventory management, is in the area of over, short and damaged (OSD) product in picking operations.
"I always look at a company's OSD to see how often they've updated information in the OSD area," says Hume, "because product that has or has not been added to the active inventory locations has a major impact on lean inventory environments."
In lean environments every piece of inventory is spoken for. When an extra piece is picked or damaged, and the system is not immediately updated—at least by the end of the shift—the next picking shift has under or over problems.
The toughest inventory management problems to solve are those created by poor IT system design. "Most often these problems revolve around capturing the wrong data set, or improper timing of updates and triggers to the inventory management system," says Hume.
He cites a case of a client receiving significant charge-backs from customers due to shipping wrong or late products. "Looking at the information flow," he says, "there appeared to be real-time inventory flow throughout the process."
Watching the process proved otherwise. At the beginning of the paint line, where product was identified as it moved in, product information was captured according to what the production plan called for—instead of what was actually being loaded. At the other end of the paint line, the data captured was of the actual output.
Occasionally, a part had to be repainted and this created all kinds of inventory,errors. On the inbound side the part being repainted was counted as a new part. On the outbound side it was counted (correctly) as the same part. The process had been designed based on the assumption that everyone was following the same plan, when, in fact, they were not.
Another issue on this paint line was improper timing of updates, Hume recalls. Triggers were set to re-order inventory when inventory levels reached a particular point. The re-order trigger was set based on the assumption of one rotation of a product. If a part had to be repainted and went around again, the trigger recognized it as part of the finished goods inventory. There was no way to recognize re-worked product. If there was any fluctuation of quality on the paint line, the operation was creating "more" product with every rotation.
In lean environments, while there are operational savings related to inventory reductions and better management, "there's a significant exposure risk from the customer's side. The drive for lean philosophies is more about internal savings for the company. Managers don't always recognize how they can potentially expose the customer to risk as a result," Hume adds.
Take a walk
"A lot of managers think evaluation of the business is going to happen in the conference room," says Connie Green, SAP business development director for Peak Technologies. "It happens down on the floor. They have to watch what people are doing. [ Inventory management is] about the least amount of movement and fewest touches to move the product through the entire supply chain."
Green has been in the business of solving inventory management problems for more than 22 years. In that time she estimates that the speed of the supply chain has tripled. "We can move stuff really fast," says Green, "but the key is to move the data just as fast. For inventory accuracy the data and the product have to move at the same speed."
Green concentrates her efforts on mobile data capture. And while going fast is good, she says you can sometimes improve inventory accuracy by slowing the process down, or capturing data at a different point of the process.
She had a client who wanted to get more product down a line and into production by speeding input. "It was capturing the data downline, manually. More stuff coming down the line faster would only slow things," she recalls. "We changed the process a bit, added hand-held data capture devices to print labels on demand, and improved the overall output because we now had real-time data information."
Green feels that making inventory improvements depends on how much vision a person has. "Managers can't always envision from where they are today, to where they need to be going," she says. "Going from a manual [data capture] system to automation has to be done in steps."
Improving inventory accuracy by going from a manual to an automated system begins with defining what " accuracy" means. "Accuracy is actually at every level of the business from ordering, through receiving and putaway, to replenishment, picking and shipping. Each area has to be addressed differently," says Green.
When forced to prioritize, accuracy at picking is the most important because picking errors lead to returns, bad customer relations and a host of other problems. "But you can't view picking in a vacuum," she's quick to note. "Putaway has to be accurate for picking to be accurate. Often picking only validates a number, not the right part in some cases."
As processes improve, inventory accuracy improves, and the flow of goods reaches a high point, will our brains be able to keep up? With a laugh Green says our brains will evolve to use exception processing.
"Probably 80% or 90% of what happens in a distribution center is repetition. Something going to a certain address," she says. "When there is a mistake, the manager's brain will register. It's those people in the back room who are going to be overwhelmed with data. The challenge will become processing the data that will give the best ROI."