December 19, 2011
technology review | Warehouse Management Systems
Implementing a new WMS across a network of DCs carries big risks and big rewards. Two shippers that are going through it tell how they're avoiding pitfalls and achieving success.
By Toby Gooley
It's hard to imagine a more daunting prospect than implementing a new warehouse management system (WMS) in more than two dozen warehouses and DCs, some company-owned and some outsourced. And it's all the more intimidating when each facility has its own approach to doing business, not to mention a unique combination of software, material handling equipment, products, and customers.
Yet more companies are doing just that. After taking a hard look at their operations, they've decided the potential benefits—accurate data, consistent service levels, and big cost reductions across an entire distribution network—outweigh the challenges and risks.
Two such companies are Kimberly-Clark Corp. and Loblaw Companies Ltd. Both are in the midst of projects to standardize their warehouse management systems across complex distribution networks. And while each has encountered a few hiccups along the way, they now have implementation down to a science. Here's a look at some of the steps they've taken to ensure success.
A "proven, repeatable implementation model"
Kimberly-Clark (K-C), the consumer and medical products titan, is in the middle of a major software upgrade—one that entails switching 21 sites over from a 12-year-old version of RedPrairie's WMS to the current version (v. 2011.2) over a period of 20 months. Of those 21 sites, 10 are field DCs run by third-party logistics companies (3PLs), and 11 are plant-attached DCs, two of which are run by 3PLs.
The facilities themselves range in size from 300,000 to 1.6 million square feet. Many were using automated storage and retrieval systems at the time of the go-live, some have multiple floors, some include co-packing operations, and, depending on the location, there can be multiple languages in use. In all, there are more than 2,000 named WMS users.
One of the first steps K-C took was to switch from local servers to a centralized server, a move aimed at simplifying programming and minimizing risk of disruption. Previously, if something went wrong, an entire facility could be out of commission while technicians tracked down the source of the problem and made repairs. In addition, code revisions had to be carried out facility by facility. With the centralized approach, that's no longer the case. "There is now a single set of code, with no unique codes at any facilities," says WMS Project Manager Bob Polar. "When there is a change, only one person approves it so we have standardization."
K-C and RedPrairie put together a corporate oversight team, with representatives from information technology (IT), distribution operations, sourcing and supply chain management, and finance. They also created multifunction "core" teams of about 10 people at each site that included distribution staff, hourly workers, and inventory control.
Once the first site was running satisfactorily, members of that DC's team went to the next implementation site to talk about their experience and offer advice. At the same time, team members from the site that was scheduled to go third came to observe and ask questions. This process will continue for all 21 sites, so that each team can learn from its predecessors and pass on accumulated knowledge to the next two down the line.
Every software implementation carries with it the risk of an operational disruption. To prevent delays in order fulfillment, the company arranged to have some of the work shifted to other DCs when the first few sites went live. But K-C was soon able to dispense with that precaution, Polar says. As the team gained experience, the need to shift volume was virtually eliminated.
On-site support has played a big role in ensuring everything goes smoothly, not just during implementation but for weeks to come. Several people from RedPrairie and Kimberly-Clark remain on-site for four to seven weeks, depending on the individual facility's needs.
Ultimately, Kimberly-Clark's aim is to have a "proven, repeatable implementation model" that results in uniform practices in all of its DCs, Polar says. "I could take someone from Paris, Texas, and ship him up to New Milford, Conn., and he could be loading trucks with the WMS in five minutes," he says. "The processes are that standardized."
A "cookie cutter" approach with flexibility
A true conglomerate, Canada's Loblaw Companies Ltd. manufactures consumer products and operates conventional and discount grocery stores under 22 banners, among other business lines. After a decade of store expansion that pushed up supply chain costs and stretched Loblaw's logistics network to the breaking point, the company in 2008 launched an initiative to redesign its warehousing, distribution, forecasting and replenishment, labor, and inventory and order management processes, supported by new software and IT infrastructure.
As part of that program, Loblaw is rolling out a new WMS from Manhattan Associates at about two dozen regional and national DCs across Canada, some company owned and some operated by third parties. Currently, 13 are running on the new WMS. The project is scheduled for completion by the end of 2012.
Loblaw has WMS implementation down to a science—it's now able to complete a rollout in 12 hours or less, including a preproduction dry run. "Early on, we diverted a lot of shipments to other DCs, but now the teams have gotten so good at this, that the last two implementations had no impact at all on operations or orders," says David Markwell, vice president information technologyâ��business solutions delivery.
Careful planning, testing, and oversight together with a formal governance structure have been key factors in the project's success. At the top is a national WMS oversight team that's responsible for activities that are "global" rather than site-specific. That team is made up of representatives from Loblaw, Manhattan Associates, and related vendors, such as the voice system provider Vocollect. Participants include functional and technical experts in WMS and labor management software, industrial engineers, data and business analysts, reporting specialists, warehouse slotting specialists, and quality-assurance experts. Among their responsibilities are system design, environment management (ensuring the new software doesn't compromise other areas' performance), code base/release management (tracking code versions and release timing to ensure consistency), and defect management.
In the second tier are three deployment teams whose responsibilities include DC-specific design and configuration, site visits and change management, preparation and training, environment preparation and setup, data conversion, readiness testing, implementation, and warranty support. The three teams work concurrently on separate implementations, each of which takes approximately 20 weeks from planning through decommissioning of the old system. This method allowed the company to go live at nine DCs in 2010.
Although Loblaw has adopted a "cookie cutter" approach to WMS implementation for the sake of consistency and speed, its process does allow for some variations in execution. These variations may be necessary to accommodate differences in building size, number of stock-keeping units, material handling configurations, magnitude of a DC's IT and facility infrastructure upgrade, and training requirements. As the project has progressed, the teams have made changes based on what they've learned. For example, the makeup of the teams evolved over time to reflect the many functions that are affected by an implementation. "We recognized that we need to approach [the project] from a holistic view," Markwell says.
Loblaw made managing risk a priority for the WMS rollout. For example, operational risk and contingency options influence the order in which warehouses are chosen to implement the new WMS. (Other factors include the line of business serviced by the DCs and alignment with other delivery programs Loblaw has under way.)
Another example is "regression testing," or running scenarios in the new software against existing functionality "to make sure everything else still does what it's supposed to do," Markwell explains. In addition, he says, the implementation teams always have a live copy of the production database, so if needed, they could shift production over to another site with minimal downtime.
For all its technological precision, the WMS project's success also depends on individuals' expertise and sense of ownership. Loblaw's regional vice presidents and the sites' general managers are deeply involved in change management and training at the facilities. Some DCs have engaged in friendly competition, vying to be the first to meet operational goals.
Markwell says a positive attitude and desire to work collaboratively are among the reasons why the WMS rollout exceeded expectations in its very first month. There were a few glitches early on, and the teams have encountered a bit of resistance here and there, but he has no doubt that there's smooth sailing ahead. "Everyone has come to understand that we should leverage this tool that we've been given to improve our performance," he says.