Process innovation at the Sandy Lumber Mill.

Metlen, Scott

CASE DESCRIPTION

The primary subject matter of this case concerns the value of quality control and continual system improvement. Secondary issues examined include project implementation, and quality management issues. The case has a difficulty level appropriate for undergraduate seniors towards the end of a semester quality management class and for graduate students. The concepts presented in the class are not trivial and several hours during class can easily be used to discuss all issues. A student's degree of understanding of process control issues, financial analysis, and quality management issues dictate the amount of time out of class each will spend to address case issues. Most students will need to spend a minimum of four hours to address all issues.

CASE SYNOPSIS

The Sandy sawmill produced dimensional lumber. The mill had recently completed a $2.6 million process upgrade that had allowed it to improve yields and better match lumber produced to market conditions. Revenues had increased from $29,000,000 to over $40,000,000 as a result. But challenges and opportunities remained. The milling process was theoretically capable of producing closer to specification, and operators still had difficulty identifying where in the process defects originated. Further, the mill manager was somewhat overwhelmed by the amount of data being produced and how best to use this data. The case ends with the mill manager wondering what he should do next to get the most out of the new system. This case was designed for use in a quality management class to facilitate discussion of the design and implementation of a statistical process control systems.

INSTRUCTORS' NOTES

Introduction

The Sandy sawmill produced dimensional lumber from a variety of soft wood species. The softwood lumber industry was a competitive commodity industry--14% of companies producing in 1997 were no longer in business by 2004. The Sandy mill had remained competitive through sound management, process redesign, and continuous process improvement, but mill manager Tony Flagor knew more could be done. In particular, the sawmill process did not recover as many units of lumber from the raw log input as possible because lumber was initially cut larger than specification to reduce the risk of producing undersized product. At the time of the case, the mill has recently completed a $2.6 million process upgrade that has allowed the mill to reduce the extent of this overcut and better match lumber produced to market conditions. This upgrade has allowed the mill to increase revenues from $29,000,000 in 2002 to over $40,000,000 in 2004. But challenges and opportunities remained because of certain systems limitations. The milling process was capable of producing closer to specification, but could go out of control easily. In addition, the quality control process was such that out of control conditions were not noticed quickly and did not indicate which part of the milling process was out of control. Further, Tony was somewhat overwhelmed by the amount of data being produced and how best to use this data. The case ends with manager Tony Flagor wondering what he should do next to get the most out of the new system and how he might approach upper management for additional investment funds to address the systems perceived limitations.

Teaching Objectives

This case was written to facilitate discussion of the design and implementation of process control systems and the impact that such systems can have on the organization. Specific teaching objectives include:

* Illustrate how improved processes and process control can provide an organization with significant cost savings in addition to enhanced quality, and illustrate how looking only at cost of quality can be used to justify investments in process control systems. The case includes numbers in it that allow students to calculate the cost of poor quality and determine the value of the process control system to the organization's bottom line, and then compare this to the cost of implementing the new system. There is also enough information in the case to determine total processes improvement return.

* Emphasize the importance of understanding inherent process variation and the ability to distinguish between product specifications and the natural tolerance limits of the process.

* Provide a context for discussing a number of issues related to the design of a control system, including what level of automated technology is appropriate, what and how much data to collect and store, how access to data is provided to those who need it, and how the design of the system impacts subsequent process improvement activities.

* Provide a basis for discussing the process of implementing new process control systems and technologies, with particular emphasis on the need to consider people issues in the implementation process. The case also provides a good vehicle for illustrating how a work design model (e.g., Hackman & Oldham's) can enhance a managers understanding of implementation issues.

* Provide a context for discussing the appropriate use of information generated from a process control system, including whether and how such information might be used to evaluate employees and managers in the organization.

Courses and Levels For Which the Case is Intended

This case was written for use in either an upper division undergraduate or MBA level course in Quality Management that includes discussion of statistical process control concepts. The case might also be used in an introductory level Operations Management course at the MBA level.

RESEARCH METHODOLOGY

The case describes a real company, real people, and a real situation, although the names of the company and manager have been changed at the request of the business. It was prepared based primarily on interviews with the company's mill manager. Some of the information for the case was also obtained through a review of some company documents and secondary research on the industry.

TEACHING SUGGESTIONS

Discussion of this case should follow four steps of cost benefit analysis: 1) what is the current situation, 2) what situation is desired, 3) how to implement or gain the desired situation, and 4) what is the benefit of gaining the new situation. To adequately address the first discussion question, students have to understand the state of the milling process in 2002, the state of that process after alterations, how the new milling and QC systems were implemented, and how much money the company saved by going to the new milling and QC systems. Question #2 asks students to evaluate the new state to determine if it is optimal, or if there is still a more desirable state to achieve. The third question deals specifically with implementation, that is, the effects of implementation. Effects include, job satisfaction, job design, and employee effectiveness. The last question asks the students to go through all four steps of the cost benefit analysis given the state of the processes at the end of the case. Students should develop a quality control policy they would like to implement, explain why they want to change the current policy, determine how much it would cost to implement the policy, what the policy would save the company, and how they would go about implementing the new policy.

1. What is the value of the current milling process and automated process control system to the Sandy mill, just from the aspect of process control?

This question provides students the opportunity to consider the process capabilities of the mill and to quantify in dollars the costs of the allowed process variation both with and without the new quality control system. Appropriate process and market parameters are embedded throughout the case and allow the student to conduct a fairly thorough analysis of the situation. The necessary parameters are presented in Exhibit TN-1. Students will have to make some assumptions about an appropriate discount rate and a time horizon for their analysis. We have used 12% and 3 years in our calculations. The first step of the analysis is to calculate the cost of boards that were too big and too small produced by the pre 2002 production process. Presented in Exhibit TN-2 are steps taken and answers to calculations used to determine that the Sandy mill was losing about $6.145 million per year to off-sized boards. The information presented in Exhibit TN-3 show the steps taken to determine the loss if Tony had not installed the QC processes where the gain from overcut would only have been $668,671.79. By adding the QC process, Exhibit TN-4 shows the mill realized a gain of $1,285,105.35 from having nearly real time control of the process. By improving the QC process and achieving real time control, Exhibit TN-5 shows the mill could realize an additional gain of $1,011,955.95 or a total gain from controlling over cut on a real time bases of $2,297,061.30. Exhibit TN-6 shows that the current NPV of controlling overcut at the time of the case was just $.604 million, which is better than if no QC process had not been included in the original project where the NPV would have been a negative $.922 million. Yet if real time control of overcut had been installed with the original project, the Sandy mill could have realized a NPV for the project of $3,003,232.05.

There are several key points here to drive home with students. The first is that the limited process control capabilities, pre-2002, were costing the organization significant money relative to the new system ($1.285 million/year), and that basic cost of quality calculations provide a means of understanding the business implications of not having better processes and process control systems. The second key point to make is that the initial process automation and added QC system could be justified based entirely on quality savings (i.e., an NPV of $.604 million on an investment of $2.602 million). Additional savings are possible from labor savings, market advantages, increased speed, and the yield increases from more intelligent decisions of how to saw logs and boards. These gains are important both in emphasizing the impact that poor quality can have as well as in justifying Tony's position that the automation did not need to lead to layoffs for it to be effective. The third key point to emphasize is that the subsequent possible investment in the board route ID and automatic process control system, which provided better real time process tracking and adjusting capability, would deliver the biggest bang for the buck--an additional $2.399 million in NPV on an investment of $140,000! This is a huge payback and illustrates the potential value of having the right data and controls with which to manage the process.

2. Are mill operators utilizing the current milling and QC systems optimally?

Students should have recognized from the first question that both the milling and QC systems have had a very positive impact on the organization, and the improvements resulting from the systems have clearly justified the cost of the systems. However, the gains realized from implementing the systems do not indicate that the systems are being utilized optimally. There seem to be two significant areas where the QC system utilization is suboptimal, which in turn causes the milling system to be operated sub-optimal.

The first significant limitation of the system relates to how the data being generated is being used--the system seems to be used primarily to detect when the process has changed sufficiently as to be causing defective product rather than to detect when the process has gone out of control. The description in the case makes clear that the QC system raises red flags for the workers when measurements exceed product specifications. From a quality management standpoint, the QC system ideally should be used to indicate when the process has shifted (or gone out of control), regardless of whether this leads to out of specification product or not. It appears that some, but not all, production supervisors are sensitive to this issue, as the case mentions that some supervisors were at times using the graphical output to detect system drift and act on it prior to a computer generated alert. However, use of the system in this way appears haphazard and driven by no standard protocol. This situation provides a good opportunity to hammer home the difference for students between the ideas of specification limits and control limits, concepts that students routinely confuse (particularly undergraduate students). What is particularly powerful about this example is that students can put a dollar value on shifts in the process that remain within the specification limits--the case indicates that every 1/1000 of an inch system drift can cost the company approximately $20,302.00 per year. Students often fail to see the point in correcting an out of control process that doesn't produce defective product, in part because it is difficult to put a cost on being out of control but within specifications. This case illustrates that a shifting process, even if it still produces products entirely within specifications, has very real costs.

Two additional teaching points can also be built into the preceding discussion of control versus specification limits and the proper use of the system to understand the inherent process variation. First, the circumstances in the case allows the instructor to make the point that the quality control system needs to be used in a consistent manner. The current situation of having some supervisors using the graphical output to try to detect process shifts, with no standard protocol and no apparent statistical foundation to their decision making, could easily contribute to increased process variability. Tony clearly needs to establish a consistent, and statistically valid protocol for the use of this graphical data. And second, Tony's contemplation of whether or not to use the data in some way to evaluate the effectiveness of operations on a given shift raises an interesting question about the appropriate uses of process control data. Certainly, Deming would argue against the use of system data toward such uses, first because it runs the risk of evaluating people based on random process variation, and second because it may cause supervisors to view the process control system in a negative way. Clearly neither of these would be desirable results. However, a review of the QC system data can reveal whether the supervisor is using the QC system appropriately to manage the milling process. At least initially, Tony would want to focus any use of the data in this way on identifying developmental/training needs of each supervisor, as opposed to making a formal part of a supervisor's evaluation system to avoid a potential backlash against the QC system.

The second significant limitation of the system relates to the design of the system and what specific information is being obtained--the limitation stems from the fact that the design of the current QC system stops short of allowing easy detection of where the milling process is out of control when it does go out of control. Because boards from different saws are mixed, and because there is no system for marking or tracking individual boards so that they can be traced back to particular saws, detection of a change in the milling process requires workers to hunt for which part of the milling process (i.e., which saw) is causing the problem. This confusion entails the workers watching each sawing step individually, pulling off a number of boards from that step, and measuring the boards. All the time this is going on, the milling process is continuing to produce sub-optimal boards and costing the company money. Once the problem source is identified, the worker can then shut down the milling process and begin to try to understand why the milling process went out of control. This situation allows for a discussion of the value of thinking about problem identification at the time the process control system is being designed. One option that is available to Tony to improve this situation is the installation of additional scanners at other points in the process or the use of a system to mark boards as they come through the milling system, as well as automatically adjusting machinery at critical tasks. Discussion of these options is probably best postponed till the latter part of the case discussion.

The material in Exhibit TN-4 indicates that the new milling and QC systems are still producing $4.86 million of boards that are either too big or too small. Because mill operators are not able to tell where an off size board comes from and from continual operation of the milling system while the operator determines the process is producing off size product and fixes the problem, too many off size boards are being produced.

Finally, it is worth having a discussion about just how much data the QC system can potentially generate and whether any of this data holds unrealized value for Sandy. Clearly, the QC system has the capability of creating vast amounts of data, and one of the challenges of the automated data collection technologies is converting all of this raw data into useful information and knowledge for the firm. It is interesting to point out to students that in some sense, Tony has both too much and not enough data. The system produces up to 1920 measurements per second (one 10 foot board with a depth width measure every 1/8 of and inch), yet Tony does not have enough data in that he cannot identify where the milling process has gone out of control. The fact that so much data already exists may make it difficult for Tony to convince upper management that more money should be spent to collect additional data. In addition, it is unclear whether full value is being realized by the huge quantity of data already being collected. Instructors may find it useful to link this discussion to section 4 of the Baldrige quality award criteria on Measurement, Analysis and Knowledge Management. In one sense, measurement, analysis and knowledge management really are at the center of the challenge that Tony faces. Section 4.1.a.1 of the Baldrige award, for example, opens by asking respondents how they "select, collect, align and integrate data and information for tracking daily operations and for tracking overall organizational performance." Section 4.2 goes on to ask about how organizations "ensures the quality and availability of needed data and information for employees ..." These are exactly the questions that Tony is wrestling with, and it is worth emphasizing to students the need for managers to constantly work toward improving information management processes.

3. Evaluate how well Tony has implemented the current systems. What has been the impact of the new systems on workers jobs? How did Tony's management of the implementation of the systems account for the affect that the systems had on job design at Sandy?

This question provides a good opportunity to discuss the affect that automated process control technology can have on line employees, and how a well planned implementation can help make this change a positive one. Hackman and Oldham's (1980) work design model, which appears in some quality management textbooks, provides a good theoretical framework with which to look at how the changes implemented at the mill influenced workers. Exhibit TN-7 shows the work design model along with notes on the influence of the new milling and QC systems. One can see from this exhibit that the overall impact of the new systems is positive on workers jobs--according to the model it should have contributed to higher internal work motivation, higher job satisfaction, and higher work effectiveness.

Once this model has been discussed and students understand that the impact of the change, theoretically, should have been positive on workers, students can then be asked whether or not they think such a change will always have a positive impact 'in the real world'. From here, the discussion can segue into the critical role that implementation played in realizing the benefits of this change on workers at the Sandy mill.

The instructor might start by discussing the moderators included in the Hackman and Oldham model--specifically workers' knowledge and skills and their need for growth. In implementing the system, Tony put significant effort into providing training for the workers and into building the case for why the change was needed. These are both critical elements in the implementation of any major organizational change. They also directly impact the moderating variables in the model. Given the characteristics of the workforce at the mill, it is clear that workers initially lacked the skills to use and benefit from the system. Had Tony not provided training, the change would have almost certainly had the opposite affect on worker motivation, satisfaction and effectiveness. Additionally, students can logically conclude that workers' need for growth was probably somewhat low. In many cases, workers were second or third generation mill workers. Many probably grew up in this small rural town simply expecting to work in the mill--to learn the craft, put in their time at the mill, and live a rural lifestyle. The young people in the community with significant need for growth probably routinely left the community in search of that growth. This characteristic of the workforce would have been an added challenge for Tony's efforts to implement the job change.

Tony's solution to the implementation impediment of liking the status quo and other impediments such as fear and/or mistrust of the unknown, and fear of losing employment (McNamara, 1999) was multifaceted. Tony convinced the workers that they needed to grow in order to maintain what they had by showing them that if they wanted to continue to prosper in the small rural community and if they wanted there to be jobs for their children at the mill, that they needed to grow. The alternative to growth would be the closure of the mill due to the increasingly competitive environment that the company faced.

Other key elements of the implementation can also be raised at this point. Tony guaranteed workers that no one would be laid off as a result of the changes being made at the mill. This guarantee was likely a hard sell to Tony's managers given the competitive nature of the industry, but was critical to getting workers' support for the changes. Additionally, Tony worked hard to involve all of his workers in the design and implementation of the changes, again in an effort to achieve buy-in from the workers and to insure a system that was compatible with the capabilities of the current workforce. As suggested by McNamara (1999), Tony also kept the communication channels open about any changes to the plan and encouraged recommendations about how the plan could be improved all the way through implementation.

In addition to the Oldham model, instructors may want to utilize the thirteen guidelines to organizational change recommended by McNamara (1999) to discuss the merits of how Tony managed implementation of the new production and quality control systems. A detailed evaluation of Tony's actions in comparison to McNamara's guidelines is shown in exhibit TN-8.

4. What should Tony do next with respect to the process control system and what affect will what he does have on the milling process?

In questions one and two students should have determined the state of the current QC system and what needs to be done to correct the system. In the answer to this last question, the students have the opportunity to utilize the design for six sigma tools to determine how much it will cost to change the QC system and what the projected savings would be from the changes they recommend. Furthermore, students have the chance to address how to convince management a project is a good project and how to get employee buy-in to the additional changes. Material presented in Exhibits TN-5 and TN-6 show steps used to calculate an expected NPV when the costs of determining which of the saws a board comes from and installing automatic process adjusting devices at each of the twelve problem areas is added to the proceeding process changes. The NPV for the total project would then be $3.003 million. The quality control process should then be built around process capability. The mean overcut and the range of the overcut should be used to determine the upper and lower control limits on X-bar and R charts. The mean and range should be determined from random sampling. Once the limits of the processing system are determined, the QC system should consist of an automated system that has been programmed to scan data for statistically significant process shifts in variation and amount of overcut. When such a shift has been detected, the milling process should adjust automatically as appropriate. Mean and range should be determined randomly and graphed on control charts for benchmarking purposes. Attribute data charts could also be utilized to determine if the frequency of off sized boards was changing. Twenty minutes of production would produce about 1320 boards with 3 to 4 boards beyond plus or minus three standard deviations thus; defects per 20 minutes of production could be graphed for benchmarking purposes. Tony also needs to work to formalize how supervisors use this information to insure better and more consistent use of the quality information generated. An additional round of training seems in order for both supervisors and operators.

REFERENCES

Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA: Addison-Wesley.

McNamara, Carter, (1999). Basic Context for Organizational Change. Management Assistance Program for Nonprofits. Retrieved June 2, 2004, from http://www.mapnp.org/library/mgmnt/orgchnge.htm#anchor494556.

Scott Metlen, University of Idaho

John J. Lawrence, University of Idaho

Exhibit TN-1: Process Parameters Needed to Calculate NPV of Changes
to the Milling and Quality Control Systems from the Perspective of
Improved Process Control

Process Parameters

 Width Depth

 0.35 0.21 inches of average overcut pre milling
 system conversion

 0.11 0.075 standard deviation in inches of overcut pre
 milling system conversion

 0.006 0.006 average inches of overcut needed so lumber
 will shrink to correct depth and width

 $0.347 $0.35 market value in dollars of one board foot
 pre system change

 $0.372 $0.372 market value in dollars of one board foot
 after system change

 0.67 0.67 number of lineal feet in one board foot
 of 2X4

85000000 85000000 board feet produced in 2002

1522388060 1522388060 lineal inches 2 X 4 produced in 2002

116,500,000 116,500,000 board feet produced post system change

 11.76% 11.76% Increase in yield due to overcut and sawing
 decisions post system change

 $0.02 $0.02 penalty per board foot in dollars if
 product is under specification

$2,600,000 $2,600,000 cost in dollars to convert milling system

$20,000.00 $20,000.00 costs in dollars for two QC scanners
 (2 $10,000 scanners)

$140,000.00 140,000.00 cost in dollars to further improve QC
 system (12 auto adjusters, ID system,
 Software)

 0.32 0.18 overcut in inches after initial milling
 system change

 0.099 0.059 standard deviation in inches of overcut
 post milling system conversion

 0.29 0.156 overcut in inches after QC1 change

 0.09 0.05 standard deviation in inches of overcut
 post QC1 implementation

 0.23 0.136 minimum overcut expected after additional
 changes in QC system

 0.057 0.026 standard deviation in inches of overcut
 after additional changes in QC
 system

 3.85 1.71 dimension of 2 X 4 after production
 pre system change

 3.82 16.8 dimension of 2 X 4 after production
 post system change pre QC1 system

 3.79 16.5 dimension of 2 X 4 after production
 post system change post QC1 system

 3.73 1.636 expected dimensions of 2 X 4 after
 production post system change and 2nd
 QCchange

 12% 12% APR or cost of capitol

 0.01 0.01 equal monthly percentage rate as decimal

 12 12 number of accrual periods within one year

 3 3 number of years project should be analyzed

 36 36 number of periods in NPV equation

30.10750504 30.10750504 PV multiplier

Exhibit TN-2: Costs of Being Under and Over Sized Pre Milling Process
Conversion

Width Depth

3.127272727 2.72 z value from .006 overcut to mean
 overcut of .35 & .21 overcut with SD
 equals .11 and .075
0.000882166 0.003264148 proportion production under the normal
 curve that would be too small
$1,499.68 $5,549.05 cost in dollars incurred for boards
 being too small
0.998235668 0.993471705 proportion of area where production
 will average .35 & .21 overcut after
 taking equal tails of distribution off
0.694 0.414 upper bound for area with average size
 of .35 & .210
0.0002394 0.0009927 conversion to determine average size of
 a given area under the normal
 curve for Z of 3.1273 and 2.72
0.694026334 0.414074453 average overcut of those boards over
 .694 & .414 overcut
0.344 0.204 average overcut sans shrinkage of .998
 and .9935 proportion of production
0.688026334 0.408074453 average overcut sans shrinkage of
 .000882 and .00326 proportion of
 production
14001742.24 3687989.468 extra board feet per year in boards due
 to overcut
$4,858,604.56 $1,279,732.35 cost to overcut per year pre system
 change.
$6,145,385.64 total cost per year of variation in
 board dimensions pre system change

Exhibit TN-3: Costs of Being Under and Over Sized Post Milling
Process Conversion but pre fist scanners for QC system

Width Depth

3.171717172 2.949152542 z value from .006 to mean of .32 & .18
 overcut with SD of .099 & .059
0.000757768 0.001593302 proportion production under the normal
 curve that would be too small
1288.2053 2708.6134 cost in dollars incurred for boards
 being too small
0.998484464 0.996813396 proportion of area where production
 will average .32 & .18 overcut after
 taking equal tails of distribution off
0.634 0.354 upper bound for area with average size
 of .32 & .18
0.0002068 0.0004553 conversion to determine average size of
 a given area under the normal curve for
 Z of 3.17172 and 2.9492
0.634020473 0.354026863 average overcut of those boards over
 .634 & .354 overcut
0.314 0.174 average overcut sans shrinkage of .9985
 and .9968 proportion of production
0.628020473 0.348026863 average overcut sans shrinkage of
 .000758 and .00159 proportion of
 production
12681070.28 3090448.521 extra board feet per year in boards due
 to overcut
$4,400,331.39 $1,072,385.64 cost per year to overcut pre system
 change.
$5,476,713.84 total cost per year of variation in
 board dimensions post system change but
 pre first scanners for QC system

Exhibit TN-4: Costs of Being Under and Over Sized Post Milling
and Scanner Installation for Quality Control Process

3.155555556 3 z value from .006 to mean of .29 & .156
 overcut with SD of .09 & .05
0.00080103 0.001349967 proportion production under the normal
 curve that would be too small
$1,361.75 $2,294.94 cost in dollars incurred for boards being
 too small
0.99839794 0.997300066 proportion of area where production will
 average .29 & .156 overcut after taking
 equal tails of distribution off
0.574 0.306 upper bound for area with average size of
 .29 & .156
0.0002145 0.0003819 conversion to determine average size of a
 given area under the normal curve for Z
 of 3.1556 and 3
0.574019305 0.306019095 average overcut of those boards over .574
 & .306 overcut
0.284 0.15 average overcut sans shrinkage of .9984
 and .9973 proportion of production
0.568019305 0.300019095 average overcut sans shrinkage of .000801
 and .00135 proportion of production
11379428.48 2616604.927 extra board feet per year in boards due
 to overcut
$3,948,661.68 $907,961.91 cost per year to overcut pre system
 change.
$4,860,280.29 total cost per year of variation in board
 dimensions post system change and first
 scanners for QC system

Exhibit TN-5: Expected costs per year of Being Under
and Over Sized Post total Quality Control Process

2.98245614 2.923076923 z value from .006 to mean of .23 & .136
 overcut with SD of .057 & .026
0.001429797 0.001733022 proportion production under the normal
 curve that would be too small
$2,430.65 $2,946.14 cost in dollars incurred for boards being
 too small
0.997140406 99.65% proportion of area where production will
 average .23 & .136 overcut after taking
 equal tails of distribution off
0.4 0.212 upper bound for area with average size of
 .23 & .136
0.0004099 0.0005053 conversion to determine average size of
 a given area under the normal curve for Z
 of 2.9825 and 2.9231
0.400023364 0.136013138 average overcut of those boards over .400
 & .212 overcut
0.224 0.13 average overcut sans shrinkage of .9971
 and .9965 proportion of production
0.394023364 0.130013138 average overcut sans shrinkage of .00143
 and .001733 proportion of production
8830190.481 2244586.311 extra board feet per year in boards due
 to overcut
$3,064,076.10 $778,871.45 cost per year to overcut pre system
 change.
$3,848,324.34 expected total cost per year of variation
 in board dimensions post system change and
 total QC system implementation

EXHIBIT TN-6 NPV for Overcut Control and from all other Sources

 Cost of waste Cost of waste Cost of waste
 pre change post change if post QC1
 had not added change
 QC1

 $6,145,385.64 $5,476,713.84 $4,860,280.29

 Relative to Relative to pre Cost of System
 preceding system change Change
 scenario

PMT $55,722.65 $55,722.65 $2,600,000.00
project
PMT $51,369.46 $107,092.11 $20,000.00
project + QC1
PMT project $84,329.66 $191,421.77 $140,000.00
+QC1&2

 Expected cost
 of waste post
 proposed QC2
 change

 $3,848,324.34

 NPV for just NPV relative Increased
 project to pre change Yield MMBF

PMT $($922,330.05) $($922,330.05) 1927.008049
project
PMT $1,526,606.36 $604,276.31 3703.473625
project + QC1
PMT project $2,398,955.73 $3,003,232.05 6619.773188
+QC1&2
Yearly increase from rest of project
Yearly increase form yield savings from $1,285,105
controlled and reduced overcut $5.477M -
$4.860M.
Yearly increase from market gain on yield $92,586.84
increase from overcut ($.025/board ft. *
3703474)
Yearly increase from ability to be flexible $2,125,000.00
relative to the market ($.025/board ft *
85000000)
Yearly increase from increased speed of milling $7831139.40
+ market gain
 (30500000 * $.372/board ft * .6683). Total
 increase is 116500-85000/85000 = 35.47%
 11.75% from yield increase and 23.71% from
 speed increase so 66.83% of the total
 increase is from speed increase and 33.17%
 from yield increase.
Yearly yield increase from increased $2,509,168
effectiveness of log and board sawing
decisions.
 Of the 33.17% increase from yield increase
 63.393% was due to cutting decisions.
 ($.372/board ft. * .3317 * .6339 * 30500000)
Yearly increase from labor savings from one $100,000.00
less employee
 Total Yearly Gain $1,394,300.00
Went from yearly income of $29495000.00 to $13,943,000.00
$43438000.00 for a total yearly gain of
($43438000.00 - $29842000.00)
Total NPV for all projects would be
$32362411.89. Equals $13943000.00/12 times PV
multiplier from TN-1 minus costs of $2.6
million and $140000.00.

Exhibit TN-7: Impact of Process Control System on Job Design
Using the Hackman and Oldham Job Design Model

 Core Job Critical Psychological
 Characteristics States

Skill variety increase, as Skill Variety
new jobs require computer
literacy, greater
troubleshooting ability,
and statistical thinking.

Task identity increases, Task Identity Skill Variety, Task
as the new jobs are about Identity and Task
managing the entire Significance combine
process, to form 'Experienced
as opposed to the previous meaningfulness of
focus on individual boards. work'. This construct
 increased.

Task significance Task
increases, as the new Significance
focus is on wood flow,
which workers come to
realize determines
long-run viability of the
mill.

Little impact on autonomy Autonomy Experienced
 responsibility
 for outcomes of the
 work. This construct
 remained unchanged..

Feedback increases, as Feedback. Knowledge of the
continuous data collection actual results of the
provides opportunity for work activity.
real time feedback on wood This construct
conversionrate that increased
contributes to company
profits.

 Outcomes of Three
 Preceding Constructs on
 Constructs Below

Skill variety increase, as
new jobs require computer
literacy, greater
troubleshooting ability,
and statistical thinking.

Task identity increases, High internal work
as the new jobs are about motivation
managing the entire
process,
as opposed to the previous
focus on individual boards.

Task significance High growth satisfaction
increases, as the new High general job
focus is on wood flow, satisfaction
which workers come to High work effectiveness
realize determines
long-run viability of the
mill.

Little impact on autonomy All four of the above
 constructs increased as a
 result of the new process
 control system

Feedback increases, as
continuous data collection
provides opportunity for
real time feedback on wood
conversionrate that
contributes to company
profits.

Relationships moderated by workers skills, knowledge, and need for
growth

Exhibit TN-8: An Evaluation of Tony's Approach to Implementation
Using McNamara's Guidelines for System Change

Guideline: Hire consultants

Action by Tony: Tony did not employ independent
 consultants. He did consult with upper
 management and some industry experts,
 and got help from his suppliers.

Result: Consultation with upper management
 smoothed the transition to new system.
 There is no evidence that the actual
 implementation process was compromised by not
 employing consultants, but it is possible
 that some of the system design challenges Tony
 is facing may have been avoided had a
 consultant been employed at the design phase of
 the project.

Guideline: Communicate need to change

Action by Tony: Tony worked closely with employees and
 suppliers. He built a strong case for the need to
 change, based on the highly competitive
 industry situation, the strong timber tradition in
 the community, and the employees' desires to live
 and work in the small, rural town in which the
 mill was located. He then presented this case
 to all employees at a company-wide meeting.
 Then Tony held numerous smaller meetings with
 his employees until he convinced everyone that
 the current system was inadequate.

Result: Employees overcame their reservations about
 change and assumed ownership of the new
 system.

Guideline: Encourage feedback from employees

Action by Tony: Tony asked employees how they thought they
 should proceed, and required employee input
 throughout the process. The case indicates
 that this was not typical for the company.

Result: Employee involvement led to employee ownership
 of the new system. Tony's breaking with
 company culture and asking for employee
 input contributed to employees willingness
 to trust Tony and embrace the necessary changes

Guideline: Know the goals of the project and stay with
 those goals

Action by Tony: Tony provided very specific project goals:
 to reduce the overcut by .06" to increase the
 board feet of lumber recovered to log board
 feet bought to 95%, and to increase the per
 thousand value a minimum of $30 per thousand
 board feet. Tony stuck with these goals
 throughout the project.

Result: Employees had a clear understanding of what
 needed to be accomplished. Sticking with
 these goals kept the process on track and
 helped drive system changes.

Guideline: Plan the change so goals are reached
 with a designated person in charge of the project

Action by Tony: Change was initiated planned and overseen by Tony.

Results: Employees experienced consistency of leadership
 over the project, providing time for them to
 develop some level of trust in that leadership.
 This trust contributed to a smooth and
 successful transition to the new system.

Guideline: Everyone knows their duties and who they
 report to

Action by Tony: The case does not provide detailed descriptions
 of the different employees, but the case
 narrative suggests the duties were well
 delineated with all employees ultimately reporting
 to Tony. The case does indicate that duties
 changed, both because the employees were
 working with new technologies and because
 there were fewer employees. Employees
 were involved in the process of job redefinition.

Result: Smooth transition--clearly defining the new
 duties (and providing guarantees that nobody
 will be laid off as a result of the change)
 reduces the ambiguity experienced by the
 employees, helping them overcome their fear
 of change.

Guidelines: Delegation of duties and responsibilities

Action by Tony: While the case does not explicitly state
 that duties were delegated to employees--there
 are numerous points in the case that do point
 to the involvement of everyone and a true team
 effort.

Results: Saw sharpening solutions and QC problems
 were found.

Guidelines: Project will take longer than planned,
 expect it

Action by Tony: Tony planned on a given amount of time to
 complete the project. There is no evidence in
 the case one way or the other on whether
 Tony expected the project to take longer.

Results: Original install did not take more time,
 but learning how to get the most out of it did.
 Tony is now concerned about returning to
 upper management to request more resources
 (with money being a bigger concern), despite
 the fact that the additional resources needed
 are relatively minimal compared to the $2.6
 million initial installation. This would seem
 to imply that Tony expected the project to
 be done the first time around, and that he might
 not have prepared his managers for follow-up
 fine-tuning of the system.

Guideline: Focus on the needs of customers

Action by Tony: The whole design was to enable responsiveness
 to meet customers changing needs
 quickly, profitably and more cost effectively.

Results: The case states that the new system "helped
 the mill become more market driven as the
 decisions on how to cut the logs and
 boards was transferred from operators to
 sophisticated computer software." This
 enabled the mill to produce those boards
 that were in greatest demand, reinforcing to
 employees and upper management the value of the
 changes made. During the implementation, it
 would have allowed Tony to justify the changes
 not just based on the long term survival of
 the facility, but also on the short and long
 term gains for the company's customers.

Guideline: Manage change, do not try to prevent it

Action by Tony: Tony embraced change, and he convinced his
 employees to do the same.

Results: Change went smoothly

Guideline: Those involved in change must take care of
 themselves

Action by Tony: Tony emphasized maintaining a healthy
 balance between work and non-work--both for
 himself (e.g., the case states he went
 biking, skiing, and spent time with his
 family to cope with the stress of the
 effort) and for his employees.

Results: Maintaining this balance helped Tony and
 his team complete a stressful project
 implementation without any real evidence of
 burnout. This should pay dividends as Tony
 moves forward--he has shown his employees
 that change can be accomplished while
 maintaining a healthy balance between work
 and non-work. This is an important step in
 motivating employees to embrace continuous
 improvement, which the case makes clear is
 critical given the competitiveness of the
 lumber industry.

Guideline: Include closure to plan

Action by Tony: The case indicates that Tony and his
 employees recognized their accomplishments, but
 also recognized some short comings of the
 system. While the case indicates that after
 quick additions to the original process,
 the team met most of their initial goals, there
 was no information provided as to how Tony
 marked closure of this part of the project.

Results: Employees have had a chance to see what they
 can accomplish, and have been able to
 take pride in their involvement. As a
 result, they may be more receptive to the
 instigation of mini projects to capitalize
 fully on the larger project. It may be
 interesting to debate with students how to
 balance "closure" with "continuous
 improvement". The challenge facing any manager
 at the end of a project is how to celebrate
 the success while at the same time preparing
 employees for the next improvement project.

Guideline: Recognize new organizational structure due
 to system change

Action by Tony: Eliminated and redefined job descriptions and
 duties to meet system requirements.
 Adjusted support systems (e.g., training)
 to match this new structure.

Results: The results were effective production due to
 leaner production. Because the new structure
 generally gave employees greater
 responsibilities, it demonstrated that
 company managers had confidence (or trust) in
 their current workforce and were willing to
 invest in them for the future. This confidence
 should build greater loyalty and trust among
 employees, which should have payback down the
 road for the firm.