Application of metrics to design and manufacture

Metrics are tools used to represent qualities which influence or result from the way things are made. Metrics are essential to the Design Process since they allow abstract ideas (goals) to be converted into measurable criteria, and the criteria allow comparison of alternative options for achieving the goals. For a characteristic such as "quality", metrics would include criteria for both fit and finish. Statistical methods used to quantify fit and finish are most often based on number of defects. Quality increases as the number of defects in fit and finish (a pair of numerical criteria) approach zero. Fit is described in metrics based on dimensional tolerances and is normally expressed in terms of an allowable range of deviation from the norm.

One essential function of the Design Process is to ensure that the overall requirements of assembly and appearance do not exceed the capabilities of the production system, resulting in the appearance of unnecessary defects. This is accomplished through the development of a product specification. The specification is based on assignment of a value or range of values to some aspect of the product which can be readily observed and which can be characterized by a measurable quality which can be easily tested.

Eli Whitney is most famous for developing the cotton gin, but one of his ideas which had even broader application was the development of "mass production". Whitney envisioned the modern mass production process, based on distributed batch production. The key was the design of a product that could be assembled from bins of interchangeable parts at assembly stations where parts were "put together" instead of "fitted together". Assembly would require assemblers with wrenches and screwdrivers, not craftsmen with files and drills and hammers.

In 1798, Whitney and Simeon North contracted with the US government, promising to deliver 10,000 muskets in 28 months. It actually took them over 10 years to fulfill the contract because the range of variability in the fit of the batch-built parts was too high to allow efficient production. The parts did not match the patterns, so either the production tolerances had to be improved or the muskets had to be crafted by hand. Years were spent reducing the range of variability of parts and adjusting the tolerance of their fit until parts could be pulled at random from piles (of graded parts) and assembled into working muskets. The statistical methods (6- or Six Sigma) devised by Whitney still provide the basic metric of modern manufacture.

Developing metrics for customer concerns

The first metric you must develop in the course of designing a new product, or revising an existing product to make it easier to build or easier to sell, is the metric for customer concerns. This must precede the development of product specifications or even basic concepts.

Norman Bell Geddes was the first industrial designer in the United States to demand that the customer's concerns be installed alongside other more traditional manufacturing concerns. The method he applied was this:

  1. Determine the precise performance requirements for the product.
  2. Study the methods and equipment used in the client's factory.
  3. Keep the design program within the budget.
  4. Consult experts on the use of materials.
  5. Study the competition.
  6. Conduct consumer-use surveys on existing products in the field.

Since the 1920's, advertising has grown into a multibillion dollar industry, and the role of marketing in product design has grown. As a result, modern practice goes much farther than Bell Geddes went. And it assigns more importance to consumer concerns. The reason for this is obvious to anyone who has gone into production: it doesn't matter how fast or how well you make something. If it isn't selling, it isn't selling.

In the 1990's you begin with a general definition of a problem (usually the apparent lack of an ideal product in some particular area) that you wish to solve (generally through the creation of an ideal product). This provides the basis for a mission statement. At any crossroads, you can refer back to this mission statement and use its guidance in the decision ahead. Once the mission statement is in hand, you can begin refining it into a product specification.

Mission statement

Product description

Business goals

Primary market

Secondary market


The purpose of market research

In the mission statement, your needs as a manufacturer are well articulated, but not the needs of your customers. Before you can develop a product specification and invest in development of this product, you must assess the needs of the customer and revise some of your assumptions (if the cost of fulfilling each assumption is not borne out by the value it provides to your company or to your customers).

You might think that you know what's going on in your business, so you might conclude that you have no need to test your own assumptions. This is an assumption that will turn out to be false. But to begin with, it is obvious that most of your assumptions need to be tested against customer's requirements.

How can you do this? You begin by collecting information from your potential product's customers and then attempt to extract from this information a definition of the potential customer's "needs". Then you attempt to organize these "needs" into a hierarchy (for example: a spectrum of concern that ranges between needs and wants) to establish a system of priorities.

This requires the development of a method for comparing the priorities of a broad spectrum of customers, because not all customers are the same or have the same needs. Not all customers are even end-users. For instance, your distributors, and even their sales staff, are in a very real way your most important customers, and their needs are often substantially different from those of the end-user.

Likewise, following the definition developed in the Japanese factories (which assumes that everyone in the production and distribution food chain is in a vendor-customer relationship), your own production workers and sales force are also customers.

From this perspective, a new concept emerges: the stakeholders. Stakeholders are the people to whom it legitimately matters how the product is designed and manufactured. The design specification must reconcile the disparate needs of these stakeholders.

Stakeholders have non-identical needs

We recruit design teams (next chapter) to assure that the specifications we develop really reflect the needs of the all stakeholders, and that we do not forget the larger issues when studying the details, nor overlook the details when studying the big questions.

A list of stakeholders in your operation might include:

If you are refining an existing product, the first problem is to identify areas in the mission statement which provide guidance. They appear in the business goals section.

If we begin by querying the production employees, we might refine the mission statement as follows:

Product description

Business goals

If we query the sales staff, we might get the following priorities:

Product description

Business goals

If all you queried were your end users, you would end up with the retail price set as low as possible, the quality as high as the materials and machinery would allow, service would be unnecessary or faultless, and you would soon go out of business.

If all you queried were your marketing people, price would be set just below the nearest competitor, and quality set just above. That makes their job easy but makes the jobs in production nearly impossible.

Developing design specifications

Development of a design specification involves assigning an appropriate value to a metric. A design specification is very much like a contract. In the design specification, all important aspects of the product are defined as clearly as necessary. The specifications for tolerances that must be maintained and the acceptable range of variability which must be maintained if the part is to function correctly in the overall design are each based on assigning values to the metric for fit.

It has been suggested that warfare provides the primary dispersant for technological innovation. Yacht designer Robert Perry speculated that the America's Cup was a charade: that if people wanted to see some real progress in the development of more efficient sailboats, they would agree to fight the next war with sail powered vessels. Lacking a "motivator" which rewarded technological innovation, the fundamental conservatism of the boat-design market (primarily driven by the needs of fishermen and freight haulers) required that progress be gradual and incremental.

Warfare does not apply only on the battlefield. The United States is often at war with much of the rest of the world economically. Time after time, we have pioneered the development of innovative products and created markets for these products, only to find our share of that market vanish as Asian firms gear up to make them better and cheaper. Better and cheaper are both powerful order winning criteria.

For years, American manufacturers attempted to foster the myth that the critical difference between a product from GE or Zenith and one from Sony or Matsushita was the cost of labor or the lack of environmental regulation in other countries. But this is not the case, and believing it is will not help us get back into the market. The difference is, and always has been, how things are made. Discussion of these differences demands the development of another kind of metrics.

Metrics allow you to "measure"

Metrics also provide the means of translating customers' requirements into specifications. Specifications are necessary tools for evaluating the success of a proposed concept in meeting the requirements of the market. Through application of other metrics to processes ranging from manufacturing to marketing, this success ought to be translatable into reproducible success in the marketplace.

The purpose of developing metrics is to provide means of expressing the characteristics of your product in terms of measurable values. This is just as important in defining your relationship with your employees or your vendors and subcontractors as it is with your customers. As you will see in the following material (and in the following section on The Role of Design), the same metrics can serve a variety of purposes. It must also be stated that the development of metrics is not necessarily the same as the development of "agreement", but is probably a necessary step toward that goal. The customer might assign one ranking one day and another ranking the next, based on issues and experiences entirely unrelated to the product itself.

In the following examples, different members of the design team use the metric table to evaluate the concept from different directions, assigning different values and "importance rankings" for the same metric.

Production looks at their own needs list and the metrics developed to assess compliance with those needs, but assigns a value based on cost to achieve that specification using available technology. Finance looks at the same metrics and assigns a different value based on their own needs and the potential to leverage the purchase of new manufacturing technology.

The basic metric matrix

Applying metrics to your manufacturing options

When you are faced with a variety of possible solutions to a problem, which one do you pick? One way to choose among a range of options is to develop "scenarios" and "play them out". This is what they do at 3:00 a.m. when they can't sleep. It also works fairly well in initial brainstorming sessions when there is nothing but the group-members' intuition to guide you. A more precise method is to apply the logic of metric-based evaluations to the decisions you have to make.

In either case, you must define the problem and develop the specification for its solution. Using these tools (metrics) from the beginning assists you in both activities and brings a reproducible level of detail and rigor not necessarily provided by other methods you may have used.

Metrics provide a very effective tool for carrying information collected in one venue to another. Information generated in one phase can provide the basis for the next level of analysis, or can be merged with information collected in other areas. In this way, the overlap between the needs of consumers, as expressed to market researchers, and the realities of available production equipment, as defined by your production manager, can be evaluated on the same page.

Developing metrics to support management decisions

Once you start watching the ratio of value earning activity to non value earning activity in your own operation, you will see how little of the time that your payroll buys is actually spent converting materials into products. This would be expressed in a metric for operational efficiency or productivity.

An old adage, often called Mickey Thompson's rule, states that speed costs money. The adage then poses the basic question: how fast do you want to go? Key to developing a speed-up strategy is assembling a collection of management information. You have to know where the limiting factors in your operation are before you spend real money to speed things up. Most business data costs approximately the same to collect and organize, but some kinds of information has far more value than others.

The greatest performance increases come at the bottom of your performance range. There are many areas where you operate between 0% and 50% efficiency, meaning that over 50% of your employees' time is spent doing things that do not earn you any money. Increasing the efficiency of these activities offers the greatest return on money invested.

One of the basic premises of this manual is the recognition that it is not cost effective to make major capital investments to incrementally speed up the part of your operation that is already working effectively. If your accounting system is up to date, and if you are not actively involved in getting invoices out and taxes paid (because these steps are being handled in a timely manner), and if your accounting system is running on an old 286 powered DOS-based computer, then investing in a Pentium and a new, Windows-based accounting system can never be justified on the basis of the increase in the speed and efficiency it will provide. The places to spend money are in the areas where things ARE NOT working, not where they ARE working.

Most of your employees' time is spent simply moving material from place to place or preparing material for its transformation. You will immediately recognize that the greatest potential to increase your profitability comes from converting this "lost" time into "productive" value-earning activity. You are currently paying as much for lost time as you pay for any other time.

In a large operation, the lost time constitutes a major gold-mine, and harvesting and harnessing this stream of money, and reinvesting it into increased efficiency and increased productivity, is the goal of management.

In the Japanese model, everyone in the operation is recruited into a simultaneous quest to increase quality without reducing efficiency, and to increase efficiency without reducing quality. This is accomplished by the gradual and continuous elimination of opportunities for defects to occur.

How to identify places where money can be saved

Based on the simple metric which separates value-earning activities from non-value-earning activities, the surest way to save money is to increase the ratio without increasing the defects. A variety of methods have been developed to achieve this end, and each offers merit. However, it must be remembered that most of these methods were developed to eliminate subtle inefficiencies in large operations over large runs. Given a million-unit run, a $0.20 savings on any item, even if it's only .001% of the cost of production, could amount to a savings of $20,000 or more. When this logic is applied across the entire production process, the potential savings often amount to millions of dollars.

The cost of implementing these programs, as they are set forth in text-books and seminars is usually out of the reach of a small company making short production runs. In spite of this, there are several general threads that run through these methods which can be applied to any project at any scale.

The most inefficient activity that happens in woodworking shops is materials handling. The inefficiency peaks at two points: unloading the delivery vehicle and moving material from the storage rack into the panel saw. The most obvious way to eliminate several phases of inefficient materials handling is to contract out the major panel cutting operations, either to the panel providers or to an out-source panel cutter.

If such a facility were available in your vicinity, this decision could reduce the cost of cutting material substantially, and help you attain competitive pricing with firms located in industrial hotbeds like Pasadena, Phoenix or Tacoma. Parts would arrive precut, labeled and palletized, and in quantities which represent a readily adjustable balance between efficiency and liquidity, (whatever quantity appears to be realistic for your rate of production). The cost of these precut, palletized parts would be highly competitive with what you are now paying for uncut panels waiting to be unloaded from the delivery truck and still needing to be cut.

Cooperative communities

The decision to buy rather than make has a tendency to snowball, and once enough companies are working cooperatively, they support the development of "communities" of specialized vendors and contractors.

You can see the pattern around any big manufacturer. Communities of related businesses develop and support one another. Silicon Valley, Cupertino and Redmond provide the high-tech version of this model (communities of software subcontractors, offering specialized out-source services, such as disk-duplication, document printing, packaging, as well as building and installing machines, etc.).

During the heyday of production yacht-building in Taiwan (the early 1980's), an entire community of secondary wood-products manufacturers and other marine-related companies grew up as sub-contractors to the large boatyards. Small firms specialized in making generic products such as brass fittings, soapstone countertops, or a particular style of louvered doors.

Advantages of allowing someone else to handle the headache quickly became apparent. All that was required of the yacht-builder was a list of doors specifying material, height, width, the direction they were to open, and the date they were to be delivered. All at once, millions of time-wasting decisions were out of his hands. The sub-contractor was responsible for the aesthetic decisions such as width of rails and stiles, the location of the handle, number of hinges required, etc.

Look, by way of contrast, at what happened in Port Townsend, our local wooden boat repair and yacht building center. For many years everyone stayed in competition with one another, rather than specializing and cooperating. A few peripheral marine-dependent business sprang up like Artful Dodger and Rainshadow Interiors. But few, if any, specialized mill-work operations are supported by the platform of sales developed by the larger builders, and most of the marine related operations survive by marketing their products and labor to the larger community "outside" Port Townsend. As a result, we never reached an operational level of community, and custom production of yachts in the Northwest was never really price competitive with Taiwan, Santa Cruz or Costa Mesa, Ca. Only one high-end luxury yacht firm (Admiral Marine) continued to grow through the first half of the decade, competing in a market where price is sometimes a secondary consideration.

Reduce part damage during assembly

One of the hall-marks of a well-designed product is the way that rough edges disappear. They are covered by subsequent attachments or processes. Good designs are easy to assemble, based on fits that are easy to maintain. Designers recognize the proper use of trim as the final detail that covers the last line of gaps and ragged edges. Well machined trim can provide a strong, finished visual line while it eliminates the need to address chips in the edge it covers. Likewise, softened corners can save substantially more time than they cost to produce by relaxing tolerances, making gaps look deliberate rather than sloppy.

Instead, quality control is most easily accomplished by working from clear instructions and by building complete units, working in isolated batches. Batch size is planned to match the individual or assembly team's production in a day, so that incomplete tasks do not slip across days or shifts.

Means-ends analysis

Metrics can be as "simple" as yes or no, (value = 0 or 1, no scale ) or as "precise" as .002". In between are stepwise assessments. However, accuracy and precision are not the same, and regardless of the level of detail you assign to your scales, the most important factors (which need to be documented) establish your own confidence in the value assigned. This confidence factor must be considered because it is often multiplicative. If you add a number in which you have great confidence to a number in which you should have little confidence, what do you get?

Consider using the following simple "Means - Ends" metric to identify places in the operation where money can be saved by outsourcing (buying rather than making) parts. Note that "yes", "no" and "maybe" are often unsatisfactory values: because they do not really allow conditions to be ordered on a scale of 1 to 3.

Buy prefabricated part Make part in-house
Does this decision reduce cost to maintain inventory? NoYes
Does this decision reduce possibility of confusion / errors / rework during production? MaybeMaybe
Does this decision reduce unit cost of this part? MaybeMaybe
Does this decision reduce or eliminate redundant materials handling? YesNo
Does this decision eliminate inefficient materials flow? YesNo

Still, you must evaluate decisions based on stable criteria and develop a system that allows you to make decisions without a lot of wasted effort. The most popular system ever invented for "normalizing" disparate values is money.

Rephrasing the questions slightly allows the answers to be evaluated in terms of their impact on your bottom line.

Buy prefabricated part Make part in-house
How much does this decision reduce cost to maintain inventory? $120,000 $12,000
Cost of rework attributed to confusion / errors / rework caused during production $41,000 $165,000
Inventory + Rework Does this decision reduce unit cost of this part? $161,000 $177,000
Does this decision reduce or eliminate redundant materials handling? YesNo
Does this decision eliminate inefficient materials flow? YesNo

Metrics also provide the means of translating customers' requirements into specifications.

Applying metrics to internal decisions

Generally, the information on which we must operate is inadequate to support a detailed analysis. In the face of inadequate information, we rely on intuition. Often, however, it is beneficial to structure our lack of information to segregate what we actually know from what we presume to know, and from what we do not know. In the course of developing material to support a decision, we are often faced with estimates of the cost of following one path or another. How can we assess their validity? How can we assure that they are as accurate as possible?

In photography, when faced with a situation where the information is inadequate or inconclusive, you "bracket". In cost-estimating, you create ranges for numbers in which you have less than the desired amount of confidence. The breadth of the range assigned to a value is a measure of confidence and allows you to apply an appropriate "discount" at a glance. This method is much safer than using multiple decimal places to indicate accuracy.

Here is another approach to developing support information for a decision. This is a representation of the kind of process used by Alderberry's owners to determine that Alderberry could not afford to continue to apply finish to the RTA furniture it sold to RainBarrel. This is a variation on the standard "make-buy equation". Simply put, you ask if consumers are willing to pay enough more for "finished" goods to cover the additional expense of producing them? Can we figure that out without investing thousands in a spray booth and new air filtration system? Finish operations are clearly problematic. If applying finish is really required to gain increased market penetration, how "much" finish application is required?

Before we can answer this question, we need to know what it actually costs to apply finish. In particular, we need to know what percentage of the perceived value of the piece and of the total cost of the piece the finish operations represent:

Assumptions are that the finish process is soya-alkyd polymerized oil finish, that drying time between coats is overnight and that it takes 3 coats to finish the pieces. This is another variation on the PDM developed previously.

Cost to apply Labor TimeEnergy OverheadConsumables Disposables
Assemble materials 4 minutes
Prep space4 minutes per unit.

assumes 5 units per batch,

20 minutes space prep

800 sf out of 6300 total tack rags

cotton gloves

Prep parts4 minutes


15 minutes per batch

Hang parts4 minutes
Mix1 minute


5 minutes per batch

Stain, solvent, oil, cobalt napthenate (drier) $28.00 paper pail, strainer, stirring wand,
Apply coat 110 minutes assumes

1 hrs per batch

wiping rags

latex gloves, cleanup solvent, respirator cartridges

Dry8 - 10 hours heat cost - 4.5 kw/h
Unhang4 minutes
Prep parts4 minutes
Hang parts4 minutes
Reapply coat 210 minutes wiping rags

latex gloves, cleanup solvent, respirator cartridges

Dry8 -10 hrs heat cost = 4.5 kw/h
Unhang4 min
Prep parts4 min
Reapply coat 310 min wiping rags

latex gloves, cleanup solvent , respirator cartridges

Rehang4 min
Dry8 -10 hrs heat cost = 4.5 kw/h
Clean up space2 min
Disposal cost
Distribute10 min
TOTAL COST80 minutes 13,500 kwh

This does not answer the fundamental question, about the "value" added by application of finish, but it allows us to discuss several other issues.

Once this level of information is in hand, other decisions can be supported as well. For instance, if the cost of applying each coat of finish is found to be approximately equal, evaluating the cost advantage of switching from a lower cost 3-coat material to a higher cost coating system that requires only two coats, or to working directly from pre-finished panels (manufactured or outworked), becomes possible. Analysis can be based on comparisons of multiple variables including time, cost of application equipment, cost of materials, (consumables and Disposables) as well as on any differences in the cost of protective gear, masking, materials handling, air circulation, prep time, etc. If, for instance, the total labor time is 3 shifts of overhead, and 80 minutes of labor, and this can be reduced by one third (one full shift), 450 kwh of electrical power, and 25 minutes of labor, this cost can be calculated, and the benefits assessed.

Developing this kind of material out of thin air takes a lot of effort, and it can be hard to justify the expense when all you are doing is evaluating one small decision. The natural temptation is to "wing it". It is equally hard to justify collecting detailed information when preparing a cost estimate for a bid. The important thing to remember is that once this information is available, it will find many uses you cannot now imagine. I contend that without this level of decision making support, you cannot hope to make the sort of decisions that good management demands.

Background on developing metrics

Design is the fundamental activity of Engineering

(Knuth 1971)

Likewise, representation is the fundamental process of design. The end product of engineering is the artifact. The characteristics of the artifact are its attributes. The end product of design is a description -- a definitive, inclusive list of the attributes of the artifact. We call this representation a Design Specification. The specification can be used to evaluate concepts and artifacts alike. From comparison of the specification with the concept, or with an existing artifact, we can determine whether or not it satisfies the objectives of the design.

Representation of the need and the process are both based on abstraction of function. Abstraction allows the discussion of functions to be separated from the discussion of other aspects of the artifact, such as structure or appearance.

Design is the process of developing a description or representation of something that does not yet exist, in order that we may discuss it and make it happen.

Artifact means, literally make (or do) art. The art we do is to organize objects into structures that provide functions. Many complex artifacts, such as the computer at which I am typing, are organizations of numerous smaller objects, each also an artifact. The most important attribute of a complex artifact is its organization. Generally we call this organization architecture. The architecture of a complex product is critical to its implementation. Product architecture became a major focus of attention in the field of computer design long before single microprocessor chips became more complex in their organization than mid-sized cities.

Computers and computer software are the most complex products our society has ever developed. One of the things we have been forced to recognize in the course of designing these projects is that the differences between an organization of procedures (a process) and an organization of individuals (a team) and an organization of linkages (a mechanism) are less important than their similarities. They are all artifacts of our effort and understanding.

Design, as a formal discipline, has blossomed slowly in the western world. Over the past 100 years, however, a great change has begun. And during the past 20 years the scope of "design" has been broadened to include more and more of the processes involved in the creation of artifacts.

Developing a new or revised product, or a new or revised manufacturing process, or a new or revised organization, can be broken down into a process or into a series of tasks or steps based on a pattern of decision-making. This process is the basis for the next session of this course.

One effect of this broadened scope has been a substantial reduction in the cost of manufacture in many fields. Another has been the development of new paths through the design process. Because the discipline of design (and the information developed through the study of design and manufacturing processes) has not been applied at all scales, manufacturing cost reductions have most often been localized in mid-sized companies. This has contributed to a stratifying effect on the manufacturing economy in general.

The functions of work orders

Some of the functions of a work order system are very similar to the cards that are used to schedule production in the Toyota JIT system. The allocation of other resources, including assembly space and tools, can be scheduled along with the withdrawal from inventory and the operator's time. In a computerized shop the flow of information can be "closed loop". In such a situation, information from past performance can be used to predict completion times and to schedule resources, and the information is kept up to date. New information from today's activity is entered into the database incrementally adjusting the systems' predictions.

It is unlikely that there will be predictable time differences between different employees' performance following essentially the same work order. If such a difference became obvious, management would create financial incentives for improvement. However, unscheduled competing demands, or bottlenecks on limited shop resources, can have a significant impact on task completion time.

To minimize predictable conflicts (collisions) on limited resources, access to these resources could be scheduled


The Management Information System (MIS) would work more efficiently if a bar-code scanner (connected to a PC which is, in turn, connected to the operation's network) were installed at each point where labeled inventory is withdrawn for assembly. Thus, the computer can assist in the check-off of all necessary parts and supplies. If this is done, the inventory withdrawal information can be automatically entered and connected to time-stamped employee activity. This is most easily accomplished if the Work Order itself is barcoded as well. The sequence might look like this: scan operation icon attached to inventory station, scan work order icon, scan employee icon, scan individual part labels.


The cost of computer hardware is always minor compared to the cost of getting it to work. The cost of barcode hardware is small since you can connect the scanner to an obsolete PC. The barcode software is also cheap (<$400 for all the software to design and print labels on your printer and to collect information that you can read into other programs). You will pay <$200 for the barcode wand or <$300 for scanning gun hardware that you can also connect to an obsolete machine (<$400). It could probably even be set up on a 286 level machine, which are essentially free these days.

You could mount the PC in a dustproof box and equip it with a touchscreen display (used in ATM's, point of sale terminals, information kiosks, and especially in restaurants and bars.) Touch screen technology requires no keyboard for entry, but requires programming the software and connecting it to another system.

For a lot less money than touch screens, you could use a formed polyethylene keyboard protector.

Getting data to migrate from the bar-code machine into your main project management, payroll, inventory and accounting systems, automatically, is another matter. At this point in the discussion, it would be nice to be able to describe a software package that you could buy for under $400 and learn in a week, that would provide all these functions and all the other functions you need, and still run in a cheap PC with under 8MB of RAM.

Fortunately, it is not necessary to install such a system to begin collecting useful information. This is fortunate because I am not aware of such a system, and do not believe that one exists at present and doubt that it will in the immediate future. Despite major progress that has been made in the development of "enabler" software over the past few years, we still have a long way to go.

Before you can fly you have to learn to walk. I believe that it is safe to state that before you can't afford not to have such a system in your operation, the tools to build it will be available, you will have information to put into it, and you will be able to afford it.

Database objects

The general family of tools you need to implement an automated, entity-wide, networked information management system is known as "Client-Server" software.

The client-server model is similar to the manufacturing model discussed in the first section. Everyone (be they another computer, a computer controlled machine, or an employee) is the client of the machine, person or process (the server) that provides the service to them.

Modern software has begun growing the ability to connect applications and pass information between them, based on an emerging set of standards known as the Open Data Base Connectivity (ODBC) Specification. This ability to connect programs is known as "interoperability" and it is a very important concept. Almost everyone who is developing serious database software tools recognizes the importance of interoperability, and as a result, the tools you will need to implement a Client-Server system are arriving rapidly. One of the roles the MTC can serve is to function as an evaluation test bed for these products as they are completed.

In the "old days" interoperability was not even smoke curling out from under the door of the advertising department. You bought a software package and paid someone a lot of money to set it up to try to provide the functions you could not afford to live without. This paradigm, as I have pointed out earlier, has advantages over the chaos of distributed personal computers. However, as the diversity of operations in this course indicates, the "mass market" to support the development of appropriate turn-key software to run a small, secondary wood products manufacturing business is not there. Essentially, each of you runs your business in a different way and as a result, you each need a different system!


I will continue to push data collection as the first step in computerization. The work order process (combined with inventory tracking, labeled parts and product checklists) is worth implementing, even a piece at a time. Using work orders in conjunction with time sheets (or as a replacement for them) assures, at a minimum, that all the necessary parts move through production as a unit with a minimum of effort. It also allows resource demands to be coordinated and accurate inventory information to be maintained.

With or without the scanner, adoption of work orders can provide the basis for a process which connects task scheduling with employee activity, as well as providing a link with shop resource scheduling and material withdrawals. And perhaps most important, it collects all this information on a single piece of paper