Material handling project managers are faced with a no-win proposition in these tough economic times. First they must fight for months to get their proposed project green-lighted by upper management. Once they get the go-ahead, the ROI has to be achieved by, oh, say, yesterday! But because that’s impossible, you must practice the art of compromise. How about ROI a year from go-live?
A one-year pay-back would have been unthinkable 10 years ago, but now it’s expected, according to Bob Kregel, project manager for HK Systems. That means project timeframes must shrink so the payback period can start as soon as possible.
“Customers need an immediate return on investment, otherwise they might not be in business tomorrow,” he says. “If it’s a little software upgrade, they want it done in a week; if it’s a multi-million-dollar project involving several vendors it could take 12 months.”
That’s if the project goes smoothly. But most big material handling system implementations eventually contract a malady called “Project Creep.” That happens when expectations change and those affected by the change want to tweak the design. “It may start with a defined scope and schedule,” Kregel adds, “but engineers can’t always leave well enough alone. They want to make it better. That’s good philosophically, but if you want to get it in on schedule and get it to start paying back money, project creep can delay the end date.”
Take a conveyor layout, for example. You work with your system vendor to come up with the best layout for material flow and you arrive at what everyone thinks is a good layout. Then you start the project. Then one of your engineers comes up with a better way to do it. That’s when you get into a cost/benefit tradeoff.
Will the payback of this better idea be enough to offset the project delays or do you put your foot down and argue “What we have may not be perfect, but it’s good enough and we need to get it up and running!”?
Controls have improved over the last few years, making some last-minute system adjustments more practical. System control schemes used to be engineered for months ahead of system implementation. Now, with modular technology and graphical user interfaces, you can make some changes on the fly relatively easily and quickly.
“There’s more forgiveness in controls and software today,” Kregel admits. “Now computer horsepower is so cheap you can easily add things and not have to worry about bogging down the system.”
Don Korfhage, president of APS Incorporated, says control modularity can help reduce project times by 20 percent to 30 percent.
“We’re working on a major program with Sandvik for the United States Postal Service, and this architecture cut design development time by 30 percent and costs, compared to a traditional PLC approach, by at least 40 percent to 50 percent. You don’t have to buy an eight-point I/O card if you only need four points, and you don’t have to buy a four-point card if you only need two.”
It also helps to institute troubleshooting as part of system implementation. Korfhage says APS works with the Wago I/O system, allowing them to distribute I/O and control, and with I/Gear, which integrates the I/O with a database that can store information associated with a job, whether it’s a build schedule or a recipe. “Wago provides the user with a reliable, cost effective structure for interfacing with a wide range of I/O that’s common in any material handling operation,” Korfhage explains. “I/Gear takes that structure and provides connectivity. What’s missing in many applications is a common data transport infrastructure, and you end up with a lot of point-to-point code written to connect these technologies. I/Gear simplifies integration, providing a conduit for moving data from the I/O to databases. Project time is shortened greatly by providing a front end that allows users and developers to troubleshoot connectivity problems.”
Adds Bill Lydon, product manager with Wago: “We came up with this concept of slice I/O. This module has a smart chip in it, and as you link these together, you form a data bus. We have over 100 types of these modules, and you select a combination of these to meet your application’s needs. With a bus coupler we tie this distributed I/O into a computer. You can get a system up very fast, buying only what you need, which is a cost savings and a panel space savings.”
A disciplined approach to project management is more desirable than ad hoc stabs. Some integrators have the process down to a science — to the point where project management itself can be modular.
“We have tried to turn project management into more of a cookie cutter approach,” says Ken Thouvenot, vice president of project management and marketing for Alvey, an FKI Logistex company. “Five years ago we started to develop a solid, repeatable project execution process. We documented it, and we try to run every project we do through that process. We’ve broken the process down into stages: quotation, contract negotiation, design, production, manufacturing, installation and commissioning. There are defined deliverables for each stage. That allows us to get repeatable output as we do our projects. That has significantly improved our success rate.”
A risk assessment and management program is key to this modular project management approach. Early in the quotation stage, technical, financial and commercial risks are identified, then an abatement plan is developed. This ensures you won’t get burned in any of those areas.
“You need to make this part of your execution process and, at least every month, go back to your assessment plan and identify new risks,” Thouvenot suggests. “Managing risks is a critical piece of the puzzle.”
The best way to do that is to clean up your act. If you’re automating poorly thought-out processes, you’ll get poorly operating automated processes. John Hill, principal of Esync, suggests you determine the value of your proposed improvements to the other stakeholders in your company. Start with some what-if scenarios. What if I asked my manufacturing plant to put bar codes on the pallets they shipped into the DC? What if I equip my pallet rack and my lift truck drivers with bar code scanners and radio frequency data collection (RFDC) equipment? What if I had a WMS that, upon receipt of the information on the inbound pallet, immediately tasked the driver to drop off that pallet or store it in a particular location?
To find the answers to these questions, you should consult various people: the lift truck operators, sales and marketing, finance, customer service and manufacturing. Find out what operational improvements in material handling would mean to their niche of the business. They might not all share in your expected benefits.
“If I improve performance in one area, what’s it doing to my performance in another?” Hill suggests you be able to answer this question before proceeding with a project. Then, if you do proceed, make sure your project team has the right people on it, and if possible, keep them involved in your next project. Hill cites the case of a WMS rollout done by Owens & Minor, the nation’s largest distributor of national, name brand medical/surgical supplies.
“They rolled out a WMS to 38 locations in 14 months,” Hill notes. “It was a cookie-cutter approach, and some of the sites were pretty tiny — 30,000 to 40,000 square feet. Others were quite large. To make this work, they had a core team and representatives from a couple of the other DCs involved in the first one. Then those people brought the team to their warehouses for the succeeding rollouts. They kept rolling it through so they always had the involvement of experienced people who would be responsible for the next rollout.”
This cookie-cutter approach helps ensure that you’re following a proven process and solid layouts. In fact, just preparing for that starting point can be beneficial.
“Often that low-hanging fruit produces great dividends and quick returns that tend to build confidence in the whole process of moving forward,” Hill says. “Even little things like incremental implementation of bar code reading and RF to get some success builds ownership among the employees who will be using this stuff when the system goes in.”
Incremental system development can cost more and take more time than a “big-bang” approach to system implementation, but it carries less risk. That’s why Frank Cameon, WMS practice leader at Esync, isn’t a big fan of the big-bang theory.
“I generally tend to lean toward the multiple-phase approach,” he says. “By breaking the project into multiple phases it allows better management control. In big-bang, you forego that control. Some people, instead of going live completely, do a parallel test before converting to a new system. Others would rather allocate more time to integration testing prior to converting to a new system. Some would rather start up with piecemeal functionality inbound, then rollout with outbound.”
He adds that a big-bang approach could work in the cookie-cutter model, once you get past the first project. That’s the project that teaches you valuable lessons and helps you find where the bugs are. Only after you get that one under your belt should you start thinking about weekend transitions from the old to the new. “If your first two DCs were 100,000 square feet, then you roll into a third DC that’s maybe 350,000 square feet, then you could probably convert over a weekend and do it well,” Cameon concludes. “But with larger DCs, there’s the challenge of training. It’s one thing to do a full, complete cut-over and go live, but it’s another to effectively train and deploy the solution to your workforce.”
Indeed, poor planning and training are project killers. So are these:
• False expectations. Will the vendor really handle a lot of the responsibilities that should be yours?
• Poor follow-through. You need the ability to deploy a project plan — to create it and manage to it without digression.
• No executive involvement. The key stake-holders often don’t find out about a project until the 11th hour. You can’t make those decisions along the way, so get executive involvement during the planning phase.
• Poor communication among cross-functional areas directly affected by the project. Keep everyone abreast of what’s coming, and when, and you’ll build a vision of the possibilities for future projects.
Avoiding these pitfalls will not only help your material handling system project succeed, it will help you develop a system of checks and balances that will ensure the payback potential of future projects. MHM
Ford’s “Common” Denominator
“Common” summarizes Ford Motor Company’s approach to both product development and project management. But common is special at Ford. With the launch of its new, global four-cylinder engine, Ford is setting a good example for anyone involved in a material handling system project. By using common suppliers, parts, practices and procedures across all of its engine plants, Ford intends to cut 15 to 25 percent of its capital equipment costs alone.
“If you went into our Romeo [Michigan] engine plant and saw a turntable, you’d see the same one in our Cleveland plant,” says Greg Johnson, Ford’s V-Engine plant engineering manager. “There might be some upgrades because of lessons learned, and that’s the beauty of working with the same supplier over and over. “We take these lessons learned from program to program and constantly improve the efficiency and safety of the equipment. By doing that we also drive the costs down.”
The processes installed at Ford’s Windsor plant are also being installed at its Cleveland Engine Plant Number One, and its Dagenham and Bridgend engine plants in the United Kingdom. This accelerates the trial and error learning process.
“Because we have these three programs in the same commissioning phase, soon to go to launch, we’re finding issues we can document for future programs,” Johnson continues. “We have a process verification manual and, at the end of each of our project checkpoints, we enter the lessons we learned for that period of the program. That’s what we take to the next program. What we learn from the implementation phase is used to adjust what we call our ‘common practice engineering methods.’ That document is shared with our other engine and transmission programs.”
The thinking is, by reducing project complexity, Ford will desensitize its assembly lines to production complexity so it can run small batches and still meet customer demand.
“The beauty of commonality is it takes away all the emotion of what we’re going to do and allows us to concentrate on the process,” Johnson adds. “The level of detail we can now get into our new processes is so much better because we have common machines and strategies. In our meetings we’re now talking about things before they’re bolted down and trying to run parts.”
“Standardization is a huge competitive advantage for us,” agrees Jeanne Geary, executive manager for V-engine manufacturing engineering. “We see cost reductions not only in the equipment we purchase, but in people, time and effort. The new Cleveland plant will help reduce our response time to customers. Because we have common methods, practices, equipment and processes, we’re not relying on personalities to deliver, but on the processes.”
The new Cleveland Engine Plant operations will go live in spring of 2004.
Liz Claiborne, Inc. fashions two DCs at once
Project managers involved in Liz Claiborne’s new distribution center in Cincinnati, Ohio, were aiming to produce a facility that could sort and ship women’s apparel and related merchandise orders to 7,000-plus retail customers. Material handling systems were designed to accommodate garments on hangers, full and split cases, individual items and products which require special handling.
The result would be a system with almost three miles of conveyor, three sortation systems, 23,000 pallet rack locations, a warehouse management system and a two-level shipping area.
All that would be challenge enough for one project, but the Liz Claiborne project team needed to get two facilities up at one time to support business growth — this one, and another in Lincoln, Rhode Island, for handling jewelry.
The biggest challenge was managing the startup of one facility while testing and training on a second fully integrated automated system.
“It was tough trying to manage one team across both projects,” says Brian O’Donnell, Liz Claiborne’s director of technical operations and planning. “Ordinarily you would like to concentrate on opening one facility at a time, but due to growing business requirements, we needed to open both facilities in the same time frame.”
Liz Claiborne conducted weekly team meetings with Siemens Dematic’s installation team, the building contractor and the various sub-contractors. The sequence of project activities was displayed on a large color-coded wall chart in the project management office. The chart included a floor plan of the facilities, showing where all the sub-systems went and how they would be connected. It became a handy visual aid for the various craftspeople on site to see their progress.
The warehouse control system (WCS) in Cincinnati was engineered to distribute product from full case and residual piece-picking into shippable units by order on a wave basis. Orders are downloaded from the WMS into the WCS, which utilizes RF hand-held terminals to direct DC associates for split-case, packing and special handling functions. It also directs associates for unit-sort induction, and directs material handling sub systems, including a container sorter, unit sorters, garment-on-hanger sorters and print-and-apply systems.
The facility’s capacity is 250,000 units per day, and it is proceeding in line with the company’s expectations.
Liz Claiborne’s Lincoln, Rhode Island DC was designed to handle its entire jewelry line. It is approximately 100,000 square feet and features case conveyor, sortation, the WMS and RF pick carts. The facility is designed to process 225,000 units on a peak day and is currently exceeding the company’s expectations.
Both projects were completed on time and on budget, and, in conjunction with two previous automation projects, allow Liz Claiborne to ship over 50 percent of its total annual unit volume more accurately and productively than was possible in 1995.