Did you know that:
Supply chain software relies on modules that can be assembled into “templates” serving particular industries? But all templates are not created equal.
Templates guide users through a workflow that covers each phase of the implementation.
“As long as you adhere to the guidelines, you get a feasible system,” says i2 Technologies executive VP Pallab Chatterjee.
According to Gartner, i2 has delivered solid templates for the high-tech manufacturing and semiconductor industries. But customers in retail, process manufacturing, transportation, and high-volume distribution should expect to invest in joint development of systems to fit their needs.
Many i2 users Baseline spoke with are implementing one or two applications, rather than buying an industry template. “You really have to take it one module at a time,” says one user.
The problem with power tools is it’s easy to cut yourself?
The biggest danger newcomers face is becoming intoxicated by possibilities. “You can’t do everything at the same time, and even if you could, you can’t manage it,” says J.B. Hoyt, supply-chain director at Whirlpool Corp. and chairman of the i2 Users Group.
The i2 Supply-Chain Planner that one apparel company uses is flexible enough that the company has configured it to handle day-to-day scheduling, rather than just the advanced planning functions for which it was designed. Because it’s trying to serve two different functions, the system wound up being unnecessarily complex, the user says. Adding a separate scheduling system would probably work better.
The advanced planning software at the heart of supply chain management is inherently memory-hungry? It’s easy to exceed the capacity of your hardware.
Planning modules work best when they can load all the variables into memory at once, rather than fetching data from a database.
That memory-resident approach can stretch the software’s limits in retailing, like at Kmart. The complexity of the model grows as a multiplier of the number of products the retailer stocks, the number of locations, and the number of time periods covered. Server memory can be outstripped, easily.
You may have to cut back your ambitions by deciding to model 12 months rather than 18, for example. Or buy more hardware.
One retail member of the user group complains that i2 keeps telling him “to buy about $3 million worth of hardware,” which his company is not about to do.
Problems can be solved with less memory, by breaking them up and solving them in pieces, Chatterjee says.
You have to feed your supply chain system the right data, and it has to be clean data?
Small inaccuracies and inconsistencies present in your legacy data for years suddenly become major problems. If data isn’t formatted the way the system expects it to arrive, you don’t get the answers you’re looking for.
Defining and redesigning processes often takes longer than anyone expects?
Projects stumble because the existing logistics or planning processes have never been documented properly.
“I’ve heard from several retailers that this issue of defining processes, making sure they’re accurate, and cleaning them up can turn a one-year project into a three-year projectit really can be that bad,” says AMR analyst Randy Covill.
When LTV Copperweld brought in i2 Factory Planner, shop workers rebelled. Because they believed management used production data in the past “as a club to beat them with,” it was hard to convince them that the data the new system required them to collect really was needed to drive efficiency, says Judi Malec-DiGioia, co-founder of the user group (www.i2-usergroup.org).
Just because they’re called “modules” doesn’t necessarily mean they were built to work together?
Many “modules” in supply chain software libraries were originally separate products acquired by the vendors to round out their offerings. Without major architectural changes, these products don’t necessarily integrate together without some help. For example, i2 repeatedly attempted to build its own integration technology to connect its modules, the last effort producing an “Active Data Warehouse” to act as a central repository for data. With version 5.2 of the i2 system, the company abandoned its technology in favor of integration software from WebMethods.