Adaptive manufacturing moves in

It's all about being demand driven not forecast driven and adapting to customer requirements quickly.

By Roberto Michel, Editor at Large -- Modern Materials Handling, 9/1/2006

Real-time data capture and exchange are the foundation of emerging supply chain practices, but aren't the end game for manufacturers and their partner networks. The real trick is getting partners to coordinate demand, supply, shipment and schedules in tight synch, say supply chain practitioners and analysts.

To make that happen, manufacturers rely on different techniques that are often intertwined. Take just-in-time (JIT) and lean manufacturing. JIT focuses on frequent deliveries of small quantities of parts to assembly plants as dictated by what's being built. Lean shares this aim at keeping inventories low, but differs by adding pull-based materials management techniques such as kanban.

JIT and lean are also subsets of a broader concept called adaptive manufacturing. The adaptive approach rejects traditional reliance on standard lead times and long-range forecasts in favor of a more demand-driven approach in which the supply side quickly adapts to what customers want.

Whatever the specific practice that supports a more adaptive approach, real-time technologies are the lynchpin because they help an enterprise know the current status of operations and events. In short, you can't adapt to events in your supply chain that you, your systems, and your partners don't know about yet.

"The real-time exchange of data is critical, but I think the biggest challenge is getting the data in a format and into systems that facilitate a decision-making process," says Jeff Hurley. He's the chief operating officer for TNT Logistics North America (904-928-1400), a third-party logistics (3PL) provider that offers JIT services out of 35 material service centers. These sites, says Hurley, use electronic data interchange (EDI) to trigger shipments, but the overall process relies on several systems.

The mix of systems that support the JIT practice known as just-in-sequence (JIS) include a proprietary parts-sequencing system. That translates EDI messages into what needs to be picked to match specific vehicle models and colors being assembled. Warehouse management system (WMS) software controls inventory and labor operations for the service centers. Radio frequency (RF) enabled bar code data capture is used at the centers to verify that picking and other movements match the EDI requirements.

"When you are operating a 500,000 square foot facility that's sequencing 25 or 30 different commodities, you can't just get raw data—you need to be able to do something with that information and put it into an organized process," says Hurley. "You need to utilize that information on the service center floor to make sure you are picking those parts at the right time, putting them in the right sequence and getting them shipped on time."

JIT has been popular in the automotive industry for at least 20 years, says Hurley, ever since EDI made it possible. In more recent years, he says, better technology such as integration middleware and event management software has emerged to improve JIT practices. What's more, he says, JIT practices are spreading to other industries such as doors and windows as well as industrial equipment.

Across all verticals, the pressure is to lower costs while maintaining service in the face of increased manufacturing outsourcing. "This trend of outsourcing to China—in particular since 2001—has put extreme competitive pressure on manufacturers," says Ralph Rio, a research director with ARC Advisory Group (781-471-1000). "North American manufacturers have responded to that in many cases by adopting lean manufacturing, recognizing the need to lower costs while improving quality and service."

While lean has its own set of methods apart from JIT, the two hold aspects in common, including the need for visibility into inventory levels. Traditionally, lean used manual methods such as kanban cards to control material. JIT, meanwhile, grew up around EDI. But both seek to keep information and material flow in tight synch.

"Once you have visibility, then you can figure out how to improve things," says Rio. "In a nutshell, visibility enables fact-based decision making, rather than opinions."

In lean plants, a "pull" method of production is used in which work cell activity is triggered by orders or a drawdown on a small safety stock at a downstream operation closer to the customer. Activity at the downstream operation then signals upstream operations when to make or move inventory.

Traditionally, cards were used as kanban signals, says Rio, but cards can get lost, and they don't provide visibility beyond the next cell. Enter electronic kanban, which not only handles signaling between cells, but can provide on-screen visibility to overall materials flow, as well as allow external suppliers to access information. According to Rio, e-kanban is offered by some manufacturing execution system (MES) vendors, some enterprise resource planning (ERP) vendors, as well as vendors that specialize in lean manufacturing solutions.

"Visibility is an important concept, but it can be hard to attach a return on investment to it," says Rio. "But with e-kanban, you can have a direct, positive effect on the financial performance of a company by reducing inventory, and improving customer satisfaction through better on-time delivery."

Jim Errington, director of sales support for ERP vendor Glovia International (800-223-3799), agrees that e-kanban enhances visibility. "If I'm a production planner sitting in an office, I don't know that a kanban has actually been signaled without some sort of electronic system," Errington says.

However, says Errington, e-kanban ideally should integrate with ERP, so that enterprise transactions such as call-offs on blanket purchase orders happen automatically when triggered by kanban. The financial and costing backbone of ERP, says Errington, still needs to run, so it's better to integrate lean automation with ERP.

MES, which ARC calls collaborative production management (CPM), also tackle visibility for adaptive manufacturing, says Rio. Today's CPM systems have workflow engines, alerts, as well as key performance indicators that provide real-time performance monitoring. In a CPM study last year, ARC found that visibility, followed by the need to improve quality, where the two top drivers behind CPM purchases.

"You can have a view into what's happening now on the floor," says Rio. "So rather than the past, where you would have reports about quality trends in the past month, you can access quality trends for the last few hours."

Visibility over orders, shipments and other key events needs to extend across a network of partners, not just the plant, says Tom Comstock, senior vice president of marketing and product management with Apriso Corp. (877-246-9393). "The complication is that you need to be agile under a model where you have less direct control," he says.

Apriso's MES and supply chain execution software makes use of a business process management engine with workflow capability that can model and coordinate events within plants or between partners, says Comstock. "You can model the whole process and manage it on exception basis all the way from a third-tier materials supplier to final shipments to retailers," he says.

JIT practices can also benefit from software that models the steps that need to be taken. Tom Kozenski, warehouse product marketing leader with RedPrairie Corp. (877-733-7724), says RedPrairie bolstered its JIT support by acquiring ALTA A/S, a provider of manufacturing event management software, in January of this year.

RedPrairie's resulting build to order (BTO) solution uses icons to show the parts and sequencing process required for each customized product, then helps manage the execution. "As a build to order system, it manages the execution and works with the WMS to make sure there is enough inventory, that it's in the right place and that the picks and shipments are done accurately," says Kozenski.

While newer software applications promise to make adaptive manufacturing easier, EDI remains a work horse for JIT. Yet EDI remains challenging to deploy in an integrated manner, says P.J. Jakovljevic, an analyst with Technology Evaluation Centers (514-954-3665).

EDI messages, such as 862 shipment requests, often need to augment or replace order information in an ERP system, and as a result, not every ERP vendor has pre-built this integration, says Jakovljevic. As for Web services making EDI obsolete, that hasn't happened, he adds, because "EDI is so important that it is often a case of, 'if it ain't broken, don't fix it.'"

TNT's Hurley concludes that it's the overall decision framework that's important, though real-time data capture and EDI transmissions are key to communicating and verifying every action in the process. As Hurley puts it, "In [JIT] operations, you can assume nothing. We need to be able to react to the changing flow of work."