Lean layers leveraged

Lean manufacturing's benefits are easy to grasp: build-to-demand, produce in small lots, and use Kanban to simplify execution. But picking enterprise and production management software to support lean is not nearly as simple.

Complexity arises because lean—interpreted broadly—takes in plant intelligence and quality analysis, not just systems for Kanban-based pull. Even Kanban-based pull itself is multifaceted.

What's more, ERP, lean enterprise, and plant automation vendors all address lean. Many see a key role for manufacturing execution systems (MES) in lean.

"There are many pieces to the puzzle," says Julie Fraser, a principal with analyst firm Industry Directions, Newburyport, Mass. ERP vendors, she says, have come a long way in their support of lean. MES is used often for lean and Six Sigma efforts, and plant intelligence software plays a role in troubleshooting areas like machine downtime.

Watertown, S.D.-based Angus-Palm, manufacturer of cabs and protective structures for equipment, is deploying a Kanban-based lean system from Demand Stream to automate lean production at its Greenville, Tenn., plant. Rusty Coleman, materials manager with Angus-Palm, believes this system should help on multiple fronts.

Automation and instant visibility of Kanban execution is one major area of benefit, says Coleman. Kanban cards were used, but material managers physically scrambled to keep up with the "discovery" task of watching for cards and expediting material replenishment.

With Demand Stream's suite and integrated data collection, Coleman says, "That whole discovery phase gets eliminated. When an operation is complete, the other people in the warehouse and I know exactly what is happening on the line."

The other major benefit, he says, stems from the core engine for dynamically recalculating Kanbans, which, as lean practitioners know, signals to make or move materials within a plant, or for suppliers to deliver material in just-in-time (JIT) mode. Without software to calculate the intricacies of Kanban, lean manufacturers make do with spreadsheets.

"Dynamic Kanban is much more than a fancy reorder point system," says Coleman. "It recalculates the requirements at every point [in your lean environment], and only triggers a need to build or buy more if actual usage is occurring."

The Demand Stream solution will integrate with Angus-Palm's Fourth Shift ERP system, which is from SoftBrands, a sister unit of Demand Stream's. The integration, according to Coleman, will reconcile the smaller lots and material movement efficiencies of lean with ERP's transactional foundation.

At Angus-Palm, ERP integration keeps the enterprise transactions up-to-date without needing to log onto ERP to release materials or cut purchase orders every time Kanban executes a step. "With Demand Stream, we'll be able to reduce lot sizes while still recording all the information we need," says Coleman.

The system's underlying engine passes see to it that a replenishment signal gets translated into an ERP purchase order or a blanket order release, explains Ian Achterkirch, co-founder of Demand Stream. "Some lean efforts get into trouble when a company introduces Kanban, but it is never integrated with the inventory system," he says. "If you can't blend the two, you add overhead, because you have to do everything twice."

Attitudes toward use of IT for lean is changing. John Richardson, CEO of MES vendor CIMNET, says, "Environments that mix traditional order-based manufacturing with lean feeder lines or downstream operations are pretty common. Manufacturers need a system in place that's able to manage both."

The other reason for a growing acceptance of IT-based lean, says Richardson, is that today's systems—compared to when lean practices first caught on—are more modular, configurable, and better at communicating data to supply chain partners. "Most of the companies we deal with expect that they will do some data collection."

Richardson says data-collection points need to be well defined and limited, and that some data is best collected via interfaces to machines and controllers—but few manufacturers today reject the idea of IT support for lean. "We don't have to fight that battle anymore," says Richardson.

Tom Buffo, director of product management for enterprise suite vendor QAD, says it has a full set of lean manufacturing modules for Kanban-based pull—including functionality for calculating Kanbans, mix analysis, and load leveling. Since coming out with lean modules in the fall of 2002, 80 companies have either deployed or licensed the solution, he says.

More lean manufacturers, says Buffo, realize they need a system that not only handles lean execution, but also captures information on optimal lot sizes, product mix, and Takt times—i.e., a calculation that helps synchronize lean production with demand. Such tools, he says, properly set up lean lines and adjust them if necessary. "Purely manual methods for lean can't easily scale across multiple work cells, lines, or plants," he says.

MES vendors' ability to support single-piece flow and Kanban vary, but most MES solutions are adept at collecting detailed execution and quality-related data.

"From a quality assurance perspective, it's hard to beat an MES for supporting lean," says Industry Directions' Fraser. "MES not only has analytics [for quality initiatives], it also has dispatching and electronic work instructions."

A lean shop, says Fraser, likely uses Kanban and methods sheets rather than schedules and work instructions, but even pure lean shops "still demand the most current data on what's happening on the floor. When you have that current data electronically, there is the confidence to take JIT to the next level."

MES generates data for quality analysis and line process improvements. It's common, Richardson says, for companies practicing lean to watch metrics closely—such as overall equipment efficiency (OEE)—especially at critical operations that quickly could disrupt the pace of lean if OEE goes astray.

Richardson says CIMNET's Factelligence solution supports work-orderless Kanban, as well as traditional production mechanisms. What's more, he says, CIMNET extended its XML-based supply chain connector module to communicate lean replenishment signals to suppliers. "To be good at lean, a company instantly needs to identify material shortages or other problems out on their lines and react," he says. "But at the same time, lean isn't just about what happens inside the four walls—it's about communicating with your suppliers."