Our Sites

Developing the value network map

Value stream mapping alternative shows promise in job shops

Figure 1
This multi-product process chart shows the processing steps to make the various components that go into a specific cabinet (Part #ED1M009-32) at C.O.W. Industries. Note the magnified portion: The chart accounts for process backtracking

In the 1980s several Toyota employees were on to something. Using just pencil and paper, they developed a new tool for mapping the organization’s operations. As early as the 1960s industrial engineers had used process mapping and material flow visualization tools, but Toyota’s method was different.

Unlike its IE predecessors, this new tool helped people to identify and eliminate operational waste. It didn’t just map processes; it mapped value. The value stream map (VSM) “has become the standard for the analysis and design of overall work flows across multiple processes and factories,” according to Rother and Shook’s Learning to See, a book on VSM published by the Lean Enterprise Institute.

Still, VSM has shortcomings in certain situations, especially within the thousands of manufacturing job shops dotting the country. It can neither map multiple value streams that dynamically share resources, nor schedule and sequence orders loaded on these shared resources per their due dates. And it has trouble with complex, multi-layered bills of material (BOMs). This is because VSM originated in conveyor-paced automotive assembly lines producing a high volume of products with low (or no) variation in the product mix--an environment so unlike the high-mix, low-volume job shop.

High-mix Challenges

For years job shops have struggled with the traditional VSM. Consider C.O.W. Industries, a job shop in Columbus, Ohio. The company tracked the component flow of just one equipment cabinet (Part #ED1M009-32) consisting of 19 components. This cabinet was part of a product family that represented a significant portion of C.O.W.’s sales and production volume, which is why managers wanted to map the product’s value stream. Still, challenges abounded. Because each component, subassembly, and final product corresponds to a unique value stream, the company had to map the entire set of value streams that comprised the final product. Besides, the company found it difficult to distinguish components with unique routings from those with similar or identical routings, especially given the large number of manufactured parts that make up a cabinet.

Some have suggested that when mapping multiple value streams that merge at shared resources, companies should start with the key components and map the others only if needed. But if the shop maps only a subset of key components that comprise the final product, how will they guarantee that an entire kit of parts will be delivered to welding and assembly in the right sequence and at the right time?

In even the simplest case of two components that arrive at a work center, one of them would have to wait for the time it would take the work center to process the other component. Imagine how inaccurate the estimate of production lead-time for the final product would be if these sequencing and scheduling delays were ignored.

Moreover, the routings of several components in C.O.W.’s product exhibited a characteristic so common in job shops but unheard of in traditional VSMs: backtracking. A part may be cut, ground, bent, welded, and then ground again before assembly. VSM does not address component backtracking to certain work centers or help determine sequencing priorities at those work centers. Nor does it handle the most critical issue in a job shop: the correct allocation of capacity on capacity-constrained resources subject to due dates, not to mention the capacity wasted if consecutive setups involve highly dissimilar jobs.

Finally, VSM does not capture information about material handling between consecutive work centers. This includes moving batches from one work center to another; the frequency of those transfers; the types of material handling equipment used; travel distance and time; and how operators signal material handlers at different locations. In practice, the material handling delays between consecutive process steps actually consume a significant portion of a product’s lead-time. And if batch transfer time far exceeds cycle times at both work centers, work-in-process (WIP) inevitably will build up between them.

The Value Network Mapping Process

Because value stream mapping has obvious problems in high-mix, low-volume environments, C.O.W. shop managers used a new tool: value network mapping (VNM). The name connotes a job shop reality: Material does not flow in just one stream; instead, many streams flow to and fro through a network of machines and processes, most or all of which process more than one part.

The new approach identifies and merges value streams that are either identical or similar; captures how final assemblies, subassemblies, and individual components relate (including parent-child relationships between components); and considers all in-house value streams that constitute the product’s complete BOM. At C.O.W. Industries, the value network mapping process worked as follows:

Figure 2
This flow diagram depicts the chaotic and congested material flow logistics so common in job shops.

1. Select a product family. C.O.W. managers already had selected the product family they wanted to map--again, a family that made up a significant portion of shop revenue. But if they hadn’t, managers could have used tools such as product-mix segmentation and product-process matrices to analyze a product mix with hundreds of different job shop routings to identify product families.

2. Visualize the flow of the entire product. Shop managers produced a multi-product process chart and an assembly operations process chart for the entire product family (see Figure 1). Next they aggregated all these routings to produce a flow diagram depicting the chaotic and congested material flow logistics common in job shops (see Figure 2).

3. Plan gemba (shop floor) walks to measure the waste involved in producing the complete product. Shop managers used a flow process chart, a classic industrial engineering tool, to help record the costs and times for all operation, storage, transport, delay, and inspection steps in the flow path of any component, subassembly, or final product. This chart also complements the flow diagram, and together they provide a comprehensive “current state” of the shop floor.

4. Analyze the flow of the entire product to re-engineer the facility layout. A well-designed facility layout ensures waste-free flow of products. To get started, managers referred to the assembly operations process chart (see Figure 3). They then created a side-by-side analysis of all the value streams to identify work centers common to multiple value streams. This was done to ensure that those shared resources were “tightly scheduled” and received parts in the correct sequence so as to not hinder overall synchronized flow of all components to the welding and shipping departments. The shipping department figured into the equation because a large number of components that comprised this product’s BOM had to be sent and received as a kit to and from the plating supplier.

5. Determine if feeder cells are possible. A feeder cell is a group of dissimilar machines that can complete a family of similar or identical parts. The cell is located immediately adjacent to the work center (or cell) that will further process what comes out of it. Statistical software such as Minitab® can help determine part families with specific components having similar routings. For example, if an assembly has a number of brackets with identical routings, and all need to be welded to the same subassembly, then a feeder cell would produce these parts as kits (each kit would correspond to one welded assembly) and deliver them to an adjacent welding booth (or cell). For more information on feeder cells and their role in production flow synchronization, see The Quantum Leap ….In Speed-To-Market, by John Costanza.

As welding cells receive kits of parts from one or more feeder cells, at the point of use are kanban squares painted on the floor. Each kanban location for a complete kit of parts would display a list of components and the number of each required for the welder to fabricate a subassembly. The feeder cells would produce and deliver kits of parts directly to these squares on demand.

This avoids overproduction and excess WIP from batch processing. By producing small quantities of many different parts to make a limited number of kits, and being located close to downstream work centers that use those kits, feeder cells help to connect upstream processes to downstream processes to produce complex fabricated products efficiently and quickly.

6. Analyze the material handling. Value network mapping captures information related to material handling between machines, including travel distance and equipment used. Another classic industrial engineering tool, a material handling planning chart, can help. Handling time between consecutive work centers, or even between feeder cells, that help to produce the entire product can be mapped (see Figure 4).

7. Develop a value network map. Figure 5 shows the completed VNM. This can be used in conjunction with the Gantt chart that displays the complete schedule for producing the components, subassemblies, and final product. The timelines in the Gantt chart should correspond to the production schedule of each of the work centers that produce different components and subassemblies that comprise the final product. The start and finish time for any component or subassembly on any machine can be immediately noted because the value network map helps managers “see the whole,” which is the entire production schedule.

So What?

Admittedly, mapping job shop operations seems a bit quixotic. So many components, so many parts, and so many part families make their way through so many routings. Job shops aren’t Toyota. Why should they go to all the trouble? Because like a value stream map, a value network map identifies value and waste. However, unlike a VSM, a VNM does so by analyzing a large sample of products that together make up the majority of the job shop’s revenue and total quantity of parts produced each year.

Figure 3
This assembly operations process chart for ED1M009-32 depicts the processes all of the components go through to become a final, assembled product.

It also helps shops evaluate data accuracy and completeness in an enterprise resource planning (ERP) system. A job shop owner cannot rely solely on gemba walks and in-process work measurement, as is typical for value stream mapping. If a shop’s ERP system has incomplete or incorrect data, or the product BOMs lack the parent-child relationships inherent in fabricated products, then a VNM can help collect the data needed to update the ERP system.

The exercise helps job shops synchronize production of all the components that comprise multi-component kits delivered to assembly or subassembly work centers. By relying on finite capacity scheduling, VNM ensures that a job shop always works with a realistic and complete schedule for producing the final product.

The VNM also helps improve shop floor layout and communication. The map shows travel distance between machines, how often that distance is traversed, as well as material handling equipment used. In other words, it reveals inefficient material handling.

Finally, the map helps execute high-impact kaizens that focus on bottlenecks affecting the overall production schedule. Recall the topmost priority of any business: Maximize throughput by shipping as many orders as possible as fast as possible. This can only be achieved by reducing delays and capacity lost at the capacity-constrained resources in the facility. Simply said, cash flow equals dollars earned from shipped orders divided by the time taken to complete and ship those orders.