Is Your ERP System a Barrier to Supply Chain Excellence?

By Mark Edmondson

Thinking about upgrading your Enterprise Resource Planning (ERP) system to solve your supply chain’s materials logistics problems? You may want to learn how the market leaders manage their materials logistics before you call your software vendor.

Dell, Toyota, and Wal-Mart-- did you know that none of these productivity leaders run large-scale ERP systems? Yet their operational excellence is indisputable and has given them a distinct competitive advantage. What makes these companies market leaders, and how can you take advantage of the same approach?

In two words- process excellence. These companies and a growing number of other market leaders recognize the importance of creating a high-velocity, yet accurate materials logistics process. The Materials Requirements Planning (MRP) evolution took us down the road of computer automation, and was one of the first applications of technology to production scheduling. It promised to be the panacea for solving all manufacturing problems. Little did we know that when we arrived at the current generation- ERP- that we would still be facing daily parts shortages, shop floor disasters and end-of-month chaos. What happened to all those promises? Why has MRP/ERP failed to deliver materials to the factory floor as promised? How are leading manufacturers managing their materials logistics?

“P” stands for planning, not pushing

The concept of creating well documented business processes and a standard work model is integral to both Lean and MRP. But instead of becoming a transaction backbone that lets companies respond faster, MRP implementations often only create more bureaucracy, cycle time, and waste. As initially conceived in the 1960s, MRP was used for planning time-phased “dependent demands” for components and assemblies. Assuming the data are reasonably accurate, MRP can be a good forecasting and capacity planning tool- that is, it can provide an estimate of dependent demands in monthly, and if managed precisely, weekly intervals.

Where companies run into trouble is when they rely on MRP technology to manage the daily ordering and movement of materials between their suppliers, customers, and their own plants- a process often called Supply Chain Execution (SCE). The goal of SCE is simple – deliver the right part, at the right time, in the right quantity, to the right location. As we will see, MRP is a planning tool- not an effective SCE solution. Using MRP for SCE creates a process where purchased parts and inventory are “pushed” to their next location based on a master production schedule (MPS) - a process often called “MRP push”. Synchronizing all production operations and materials movement with a centralized MPS is an elegant idea on paper, but in reality few companies have successfully sustained the process and data integrity required to make this work.

The charm of MRP logic

As a forecasting tool, MRP is not complicated; indeed it has a certain logical charm. An MPS is used with a Bill of Material (BoM), along with assumptions for lead times and consumption rates to calculate “dependent demands” for components and assemblies. These requirements are then balanced against the aggregate of on-hand inventory, work-in-process and open purchase orders to determine net, time phased requirements.

During the 1990s, computer hardware and software advancements made it possible to create massive, fully “integrated” enterprise software packages. ERP extended the simple logic of MRP with much greater function and complexity. The holy grail of a single enterprise database was promised, integrating MRP with floor control, supply chain execution, advanced planning and scheduling, product data management, accounting, labor reporting, inventory tracking, and other systems. This was when MRP was pressed into service as an SCE solution- often creating new levels of materials logistics problems.

The unstable MRP push process

Many things have to go right for the MRP push process to work effectively as an SCE solution. The linkage between the MPS at the start of the process and delivery of an individual part at the end of the process is long and ever changing- and it responds much like when you push on a rope. The inputs to the MRP push process- the MPS, the BoM structure, engineering change levels, feature and model variations, inventory levels, outstanding purchase orders, process yields, lead times, delivery locations- all of these elements must interact flawlessly.

This process design is not only complicated, but also inherently unstable and tends to spiral out of control after any significant change. Change any of the MRP inputs and voila- for the next regeneration you have a new production schedule that is disruptively different than before. Consequently, there is little obvious relationship between the materials requirements regenerated by MRP and the reality on the factory floor.

The data monster

The challenge of working with an unstable MRP push process is magnified by its demanding data requirements. It is not unusual for a manufacturing process to involve several thousand unique part numbers. For MRP to accurately manage their requirements, all of the MRP inputs listed previously- BoM structure, engineering change levels, feature and model variations, inventory levels, outstanding purchase orders, process yields, lead times, delivery locations- must also be accurate for every part.

This is a data integrity nightmare not achievable by even the most tightly run businesses. With every data change and MRP regeneration, new short lead time dependent demands are created, causing more and more parts to be unexpectedly pushed from suppliers.

Yet despite all this planning and effort, little progress is achieved with throughput- indeed delivery performance to the customer usually suffers. The result is an MRP push process spiraling out of control. The ensuing chaos on the factory floor causes more customer shipments to be missed, which creates a new “recovery” schedule and yet another round of MRP regeneration.

What are the symptoms of a struggling MRP push process?

To see the waste of an MRP push process, just walk onto the factory floor- especially during the last weeks of the financial quarter. The symptoms that you’ll see include:

A well staffed materials “planning” or “analyzing” department
Complicated BoM structures with “phantom” assemblies and operations
“Lumpy” or “whip-saw” dependent demands
High volumes of unexpected purchase requisitions and orders
Increasing inventory levels
Continual parts shortages handled by a full-time staff of parts expeditors
Mediocre delivery performance
Lengthy production status meetings
Resources consumed with the following wasteful activities: frequent MRP regeneration and “back-flushing”, line-balancing, WIP reporting, order confirmation, kitting.

You'll find that achieving quarterly financial and production targets requires overtime, line balancing, expediting, last minute production runs, and even a little smoke and mirrors. This chaos results in scrap, rework and additional labor costs- waste that significantly impacts bottom line performance. This is hardly the automated, integrated SCE process promised by ERP. As consultants who have worked with many manufacturing companies, we’ve seen too many ERP initiatives that ultimately collapse under this spiraling burden of activity and cost. Tremendous time and millions of dollars are spent acquiring, implementing and maintaining ERP systems only to result in mountains of inventory, poor delivery performance, and frustration.[*]

So what’s the answer?

Most companies that have gained control of their SCE process have moved away from the MRP push process in favor of a simple, direct solution that uses the Lean concepts of flow, pull, and visual management. “Pull based” processes are better suited for SCE, and consistently achieve the goal of “having the right part”. Compared to technology-based MRP/ERP systems, pull based processes are inexpensive, rapidly implemented, require fewer resources to manage, significantly improve delivery performance, and enable visual management of the value stream.

Pull based SCE processes, which we call LEAN Logistics, are based on proven tools from the Toyota Production System. LEAN Logistics is not a software program, training, or pre-packaged process. It’s a SCE solution custom tailored for your business and expertly implemented with the involvement of your team.

What results can you expect?

Typical results for a LEAN Logistics solution during the first few months include:

Increased throughput: 20% - 100%
Decreased inventory: 30% - 80%
Reduced lead times: 30% - 80%
Less materials planning workload
Less materials handling workload
Improved delivery performance
Improved purchased parts quality
Better supplier relations
Lower IT costs
Increased enterprise agility

Your company’s prognosis

With a LEAN Logistics solution in place, the acquisition and delivery of purchased parts is a logical extension of your internal production process. This enables your extended value chain- your enterprise, your suppliers and your customers- to achieve industry leading agility, throughput, and efficiencies.

Who are the market leaders of tomorrow? It’s a safe bet that it will be those companies that practice operational excellence- both within their enterprise and throughout the extended value chain.

What’s the next step to learn if LEAN Logistics is right for your company?

As an initial step to working with us, we’ll schedule a 1½ day Qualification Interview. Your site leadership and key members of your team will meet with us to explore your objectives, organizational structure, major value streams, and challenges. At the end of this meeting, we’ll present a joint vision of what’s possible during your Lean Enterprise Transformation, the tangible financial and operational results that we commit to help your team achieve, a high-level timeline of events, and the project cash flow, inclusive of our fees.

Mr. Edmondson is President of LEAN Affiliates, a network of leading Lean Enterprise Transformation consultants across North America. His perspective is drawn from working with over 50 companies as a business process consultant, consulting principal, ERP and SCE solutions executive, global alliance executive, and enterprise change agent. He earned an MS and BS in Management and Industrial Engineering from Stanford University and a BA in Management and Economics from Claremont McKenna College. He also achieved certification by the American Production and Inventory Control Society (APICS CPIM).

[*] The Chaos Report published in 1995 by the Standish Group shows a staggering 31% of ERP projects are canceled before they are completed. Further results indicate that over 52% of projects cost more than 190% of their original estimates. The cost of these failures and overruns are the tip of the iceberg. The opportunity costs are not measurable, but could be trillions of dollars in the United States alone.