227 10 Process Analysis Jack B. ReVelle, Ph.D. 10.1 DEFINITIONS • Activity. A measurable happening that occurs over time. • Annotation. The process of assigning specific codes or symbols on a process flow chart or process map so as to identify the specific location where defects or errors are created, where excessive cycle time is con- sumed, where cycle time is most unpredictable, or where unacceptable costs are generated. • “As-Is” condition. The way a process or system actually functions or operates without regard to whether it is efficient, effective, or competitive. • Event. A nonmeasurable happening that occurs at a specific time, e.g., the start or finish of an activity. • Parallel Events. Two or more events that take place simultaneously, i.e., concurrently. • Parking Lot. A place or location where ideas, concepts, and suggestions for process improvement are recorded when they are conceived for easy reference at a later time, e.g., a white board or easel paper. • Predecessor Event. An event that must take place prior to the start of a specific event. • Process. A series of sequentially oriented, repeatable events having both a beginning and an end and which results in either a product (tangible) or a service (intangible). • Process Analysis. Examination of a process using tools or methods such as process flow charts, process maps, and annotation. The purposes of a process analysis are to expand the process stakeholders’ understanding of the entire process from suppliers to customers, including the critical linkages between the quality requirements and performance metrics of both inputs and outputs, and of the ways in which the voice of the customer drives the process. • Process Analysis and Improvement Network (PAIN). An integrated col- lection of process flow charts designed to facilitate understanding and enhancement of existing processes, both production and transactional. • Process Flow Chart. A one-dimensional collection of geometric figures connected by arrows to graphically describe the sequential occurrence and interrelationships of events in a process. • Process Improvement. Enhancement of an existing process by slightly improving various phases or by redesigning all or most phases. SL3003Ch10Frame Page 227 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 228 The Manufacturing Handbook of Best Practices • Process Map. A two-dimensional version of a process flow chart that also portrays handoffs and receipts of products or services from one person, organization, or location to another. • Series Events. Two or more events that take place sequentially, i.e., one following or preceding another. • “Should-Be” condition. The way a process or system should function to be most efficient, effective, or competitive. • Successor Event. An event that must take place following the finish of a specific event. • System. A collection of processes, arranged in series or parallel, that has a common beginning and a common end, and which together constitute a program, a project, or an entire organization. 10.2 PROCESS ANALYSIS 10.2.1 P ROCESS What is a process? A process is a series of sequentially oriented, repeatable events having both a beginning and an end and which results in either a product or a service. A product, of course, is something tangible, something you can see, taste, or touch. A service is something intangible, something that you can’t see, taste, or touch, but which you know you’ve received. For example, delivery of training is a service. 10.2.2 S YSTEM Well, if that is a process, then what is a system? A system is a collection of processes arranged in series or parallel, and which together constitute a program, a project, or an entire organization. A company, large, medium, or small, is an example of an entire organization. An initiative might be a project such as the initial use of some new software. A program could be an ongoing activity that is done periodically. In any case, whether it is a program, a project, or an entire enterprise, it’s a collection of processes. 10.2.3 P ROCESS F LOW C HART Having defined for baseline purposes what a process and a system are, now let’s review what we can do to better understand these processes, these basic elements or components of an organization. There are a number of different ways we can analyze a process. The most common and one of the most useful forms is a graphic tool known as a process flow chart . This chart is a series of geometric figures — rectangles, diamonds, and circles or various other shapes — arranged typically from left to right, and from top to bottom, connected by lines with arrowheads to show the flow of activity from the beginning to the end of the process. When a process is being created or an existing process is being analyzed, it is useful to create a process flow chart so that everyone involved, that is, all the stakeholders in the process, can see exactly what is supposed to happen from beginning to end without having to try to imagine it. Each of us may have a picture SL3003Ch10Frame Page 228 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC Process Analysis 229 in our own mind, a graphical portrayal of what the process flow looks like, but the reality may be different. The only way we can be sure we understand that we have a common perspective or outlook on the process is by graphing it as a process flow chart, a linear or one-dimensional process flow chart. I say one dimensional to distinguish it from the two-dimensional graphic that we are going to talk about shortly, known as a process map. Let’s talk about the creation of the process flow chart. Traditionally, people have created process flow charts from the first step to the last. I don’t, and the reason is that, when people put this flow chart together, they are looking at processes in the same way they look at them every day, so there is a high potential for missing something. What I suggest people do as we bring them together in a room to create a process flow chart is to start with the end in mind, a concept understood by everyone familiar with Stephen Covey’s The 7 Habits of Highly Effective People. We begin by defining the last step or the output of the process and then start asking the question sequentially, “What has to happen just before that?” If we know we have a specific output or step, we ask what must be the predecessor event or events that must take place to satisfy all the needs so that the step we are looking at can take place. So we work backward from the last step to the first step and keep going until someone says, “That’s where this whole thing begins.” Now we have defined, from the end to the beginning, the process, graphed as a process flow chart. Some people might question why you want to do it that way. The analogy I use that is very effective is this: suppose I were to ask you to recite the alphabet. You would say A, B, C, D, E, F, G … without thinking, because you have done it hundreds, perhaps thousands, of times. But if I were to ask you to recite the alphabet backward, you would probably say Z and have to stop and think what happens before that, what letter precedes Z. What most people do, I have discovered, is first to do it forward to find out what the letter is and then come back and say that the letter before Z is this, and the letter before that is this, and so on. Working the alphabet backward makes people look at it in a way they have never looked at it before, noticing the interrelationships between the predecessor and the successor events. The same psychology of working backward applies in dealing with our pro- cesses, whether we are dealing with a process of building a home, working with accounts payable, developing a flow chart, understanding a process as it relates to training, or whatever the case may be. Establishing the process flow chart from the last step to the first step is a very strong and powerful way to help people understand what their processes really look like. 10.2.4 P ROCESS M AP Once the process flow chart has been created and everyone is satisfied that it truly reflects the order in which the events take place with regard to predecessor and successor events, the next step is to create a process map. Earlier I said a process map is created in two dimensions. We are going to use exactly the same steps we used in the process flow chart, except now, instead of just having the flow go from left to right, we take the people, positions, departments, trades, or the functions that SL3003Ch10Frame Page 229 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 230 The Manufacturing Handbook of Best Practices are involved in the process. and list them vertically down the left-hand side from top to bottom. For example, it might be department A, B, or C; person X, Y, or Z; or trades such as concrete, plumbing, or framing. Then, we take the rectangles that we created in our process flow chart and associate them with the various functional areas, departments, persons, or trades listed on the left-hand side. What you see is a series of rectangles being built from left to right and also moving up and down the vertical axis we have created on the left-hand side of our process map. In so doing, we see what might look very much like a sawtooth effect with blocks going up, down, and across. Thus we end up with a view of the handoffs from one person to another, one function to another, or one trade to another, so we can see where queues are being built and where the potential for excess work in process is being created among the various areas of responsibility (listed down the left-hand side). This gives us a very clear, visual picture of some of the things we might want to consider doing in terms of reordering the various steps to minimize the total number of handoffs that are a part of this process, recognizing that every time there is a handoff, there is a strong potential for an error, an oversight, something left out, a buildup of a queue, the creation of a bottleneck, or the like. In creating our process map we gain tremendous insights into what we can do to continuously improve our processes. Remember, the order of the steps may have been absolutely vital at one time, but with changes in technology, people, and responsibilities, what we did then may no longer be valid, and we need to periodically assess or review our processes. The use of a process map is an excellent way to do that. Now, in addition to looking at the process flow chart and process map in terms of the sequence and the handoffs, we can also use the process flow chart and the process map to assess cycle time and value-added vs. nonvalue-added events or steps in the process. The technique I use is to ask everyone in the room to assess the cycle time of the process that was just evaluated using a process map or process flow chart. Does it take 3 hours, 5 days, 10 weeks — whatever? When we get an agreement of 6 to 8 hours or 6 to 8 weeks — whatever the final range may be — we go through and evaluate each individual step, asking how long each step takes. When we have gone all the way through that, we arrive at the grand total of all the individual step estimates and compare that to the estimate that the group has already made of the overall process. What we frequently find is that the sum of the individual steps is only 20 to 30% of the overall total. That quickly presents an image of a lot of lost and wasted time, costly time that could be used for other, important purposes. If, for example, a process is estimated to take 6 weeks, but the sum of the individual components takes a week and a half, it’s obvious that we have some time we can save the company. Now, what needs to be done? Where are the barriers, the bottlenecks in the process that we can study, where can our trades (for example) share responsi- bility? Instead of having a particular trade come back three, four, or more times to do some little job that takes a half hour, an hour, another trade already on-site could be doing it for them. That is a very effective way of reducing cycle time. Steps can be eliminated and days upon days can be banked for use in more important projects. SL3003Ch10Frame Page 230 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC Process Analysis 231 10.3 PROCESS IMPROVEMENT 10.3.1 “A S I S ” VS . “S HOULD B E ” Now let’s look at the “as-is” vs. the “should-be” conditions. When we create the first process flow chart or process map of an existing process, we refer to that as the as-is process, i.e., the status of a process as it is currently operating. It gives us a baseline to create the new, revised process that we call the should-be process. Working together, the process improvement team is now able to view the as-is process in juxtaposition with the should-be process that they have created. Subsequent to the creation of the should-be process map, the team begins to build a bridge from the as-is to the should-be process. The bridge is supported by a series of steps that we must go through to change the process from the as-is way to the way it should be. A good example of that is the creation of some superhighways where conven- tional surface roads exist. During the building effort, traffic still has to flow, so as we move from the as-is surface streets to the should-be superhighway, we have to go through a series of steps, closing down and opening various components of the roads to support as much as possible the flow of traffic that never stops. Picture the Los Angeles freeway traffic any time of the day or night. This approach graphically illustrates what we need to do to move from the as-is process map to the should-be process map. These are things we might have otherwise overlooked. 10.3.2 A NNOTATION Using either a process flow chart or a process map, a process improvement team can easily identify specific locations within a process where events should be mon- itored to determine the extent of defects, errors, oversights, omissions, etc. Moni- toring is usually accomplished using statistical control charts, e.g., X-bar and R, C, P, Np, U, and other charts. Chapter 15 on statistical process control (SPC) presents information on this topic. Annotation is the development of a listing of defects and variances associated with the process being analyzed. Each known defect or variance is assigned a number by the team. Then the team annotates (assigns) each defect or variance to one or more events on the process flow chart or map. At this point the team evaluates the combined impact of the defects or variances at each event. Based on this evaluation, the team determines where SPC control charts should be physically located on the manufacturing floor, design center, or office. In addition, the team identifies which defects or variances should be counted (attribute/discrete data) or measured (con- tinuous/variable data). The combined effect referred to above is determined by the quantity of defect or variance identification numbers annotated at each event. Those events with the greatest incidence of identification numbers have a greater need for monitoring using SPC control charts than the events with few or no identification numbers. This is a simple application of the Pareto principle, also known as the 80-20 rule. In this case, 80% of the SPC control charts will be needed to monitor 20% of the process events. SL3003Ch10Frame Page 231 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 232 The Manufacturing Handbook of Best Practices The annotation methodology is also valuable in identifying where, within an as-is process, changes are needed in the creation of a should-be process. 10.4 PROCESS ANALYSIS AND IMPROVEMENT NETWORK (PAIN) 10.4.1 R EASONS FOR PAIN There are several reasons for using the Process Analysis and Improvement Network (PAIN). Whenever a process exhibits undesirable attributes, it is incumbent upon the process owner, process stakeholders, members of a process improvement team (PIT), or any other interested parties to take timely and appropriate corrective actions to eliminate or at least to reduce the presence or influence of the negative attributes. The most common of these negative attributes are • Process too long (excessive cycle time) • Process too inconsistent (excessive variation) • Process too complex (excessive number of steps) • Process too costly (excessive cost per cycle) • Too many errors (poor quality — transactional process) • Too many defects (poor quality — manufacturing process) • Insufficient process documentation (for training or benchmarking) 10.4.2 PAIN — M AIN M ODEL (F IGURE 10.1) • Senior management identifies a process critical to success of the organi- zation. • Senior management establishes a team composed of the process owner, process stakeholders, and process subject-matter experts (SMEs). • Convene the team with a facilitator experienced in process analysis and improvement. • Have the facilitator provide a tutorial on the development of an as-is process flow chart. • Start the development of the as-is process flow chart with identification of the final step in the process and then a backward pass through the process, finishing with its first step. • Complete the development of the as-is process flow chart with at least two forward passes. • With the assistance of its facilitator, the team should now convert the as-is process flow chart into its corresponding should-be process map. • At this point, the process improvement team has a variety of options from which to select, depending upon its objectives. As noted above, there are a number of reasons for PAIN. The following models and discussions are offered to clarify the team’s choices. • When the team completes one or more of the following models, there are three steps remaining to complete the PAIN. These steps are spelled out in the final blocks of the PAIN — main model (Figure 10.1). SL3003Ch10Frame Page 232 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC Process Analysis 233 10.4.3 PAIN — M ODELS A T HROUGH G PAIN — Model A (Figure 10.2). The objective of this sequence of events is to reduce process cycle time. The process improvement tools, cause-and-effect analysis (also known as the fishbone diagram or the Ishikawa diagram), and force field analysis are explained in numerous books on continuous improvement. PAIN — Model B (Figure 10.3). The objective of this sequence of events is to reduce process variation. The process improvement tools, cause-and-effect analysis (also known as the fishbone diagram or the Ishikawa diagram), and force field analysis are explained in numerous books on continuous improvement. PAIN — Model C (Figure 10.4). The objective of this sequence of events is to reduce the number of process steps. This is accomplished primarily by identifying the value-added (VA) and non-value-added (NVA) steps that exist within the as-is process. PAIN — Model D (Figure 10.5). The objective of this sequence of events is to reduce the cost per cycle of using a process. After determining whether the costs in question are direct or indirect and the pertinent cost categories, the objective is accomplished through the sequential use of several process improvement tools. The improvement tools, Pareto analysis, cause-and-effect analysis (also known as the fishbone diagram or the Ishikawa diagram), and force field analysis are explained in numerous books on continuous improvement. PAIN — Models E and F (Figures 10.6 and 10.7). The objective of these models is to provide guidance in the reduction of transactional errors and defects (Model E, FIGURE 10.1 PAIN — main model. Process analysis and improvement navigator. ID Critical Process Establish Process A&I Team Convene Team with Facilitator Facilitator Provides A&I Tutorial Team Starts Process Flow Chart with Backward Pass Team Completes "As Is" Process Flow Chart with 2 or More Forward Passes Team Converts Process Flow Chart Into Process Map Team Creates "Should Be" Process Map Team Leader Facilitates Transition to "Should Be" Process Repeat, as needed, Biannually Team Determines Its Objectives B C D A E F G Reduce Process Cycle Time Improve Process Documentation Reduce Production Defects Reduce Transactional Errors Reduce Cost per Cycle Reduce Number of Steps Reduce Process Variation SL3003Ch10Frame Page 233 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 234 The Manufacturing Handbook of Best Practices FIGURE 10.2 Objective: reduce process cycle time. FIGURE 10.3 Objective: reduce process variation. A Measure/Estimate Individual Step Durations Calculate Overall Duration Apply 80/20 Rule to ID the Few Steps that Consume the Most Cycle Time Use Cause & Effect Analysis to ID Cause(s) of Excessive Time Consumption ID 2 or More Potential Corrective Actions Use Force Field Analysis to Select the Most Appropriate Corrective Action Collect Cycle Time Data to Confirm Cycle Time Reduction Implement Corrective Actions B Measure/Estimate Individual Step Durations(Most Optimistic Pessimistic) Collect Data When Estimates Vary Apply the 80/20 Rule to ID the Few Steps Containing the Most Variation Use Cause & Effect Analysis to ID Cause(s) of Excessive Variation ID 2 or More Potential Corrective Actions Use Force Field Analysis to Select the Most Appropriate Corrective Action Implement Corrective Actions Collect Individual Step Durations to Confirm Variation Reduction SL3003Ch10Frame Page 234 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC Process Analysis 235 Figure 10.6) as well as production errors and defects (Model F, Figure 10.7). The model is based on the Deming-Shewhart Plan-Do-Check-Act cycle. The earliest version of the model was created in 1985 as a part of a continuous improvement seminar. When the model is first introduced to a process improvement team, it is important to gain consensus from the team members regarding the rationale of event selection and arrangement. FIGURE 10.4 Objective: reduce number of steps. FIGURE 10.5 Objective: reduce cost per cycle. C Determine Desired % Reduction in No. of Steps Count Actual No. of Steps Calculate No. of Steps to be Reduced Count No. of Hand-Offs from One Function to Another ID Non-Value-Added (NVA) Steps Redraw Process with VA Steps above Time Line & NVA Steps Below Calculate VA/NVA Ratio (Bigger is Better) Set Improvement Goal for VA/NVA Ratio ID NVA Steps that can be Totally Eliminated ID NVA Steps that can be Combined with Other NVA Steps or with VA Steps Select NVA Steps for Elimination or Combination Perform Necessary Changes to Eliminate /Combine NVA Steps OR Type of Cost ? Identify Cost Categories Identify Cost Categories Direct Cost Indirect Cost Collect Relevant Recent Cost Data Failure Costs (Internal & External) Appraisal Costs Prevention Costs Conduct Pareto Analysis of Cost Data Use Cause and Effect Analysis to ID Cause(s) of Greatest Cost ID 2 or More Potential Corrective Actions Use Force Field Analysis to Select Most Appropriate Corrective Action Implement Corrective Action Collect Additional Cost Data to Confirm Cost Reduction Controllable Expenses Clean-Up Expenses Miscellaneous Expenses D SL3003Ch10Frame Page 235 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 236 The Manufacturing Handbook of Best Practices FIGURE 10.6 Objective: reduce transactional errors/defects. E Identify Internal (Operational) & External (Customer) Transactional Problems Prioritize Problems Select & Define Primary Problem Identify Process(es) Associated with Primary Problem Describe Process(es) Identify Specific Steps which Require Analysis Collect Attribute or Variable Data Determine Defects to be Counted & Charted (Performance Metrics) Determine Critical Dimensions to be Measured and Charted (Performance Metrics) Design Data Collection (Tally) Sheet Decide How Collected Data Should be Bundled & Graphed Initiate Data Collection Analyze Data For Process Competitiveness (Quality,Cost,Sch.) Process Competitive ? Institute CI Determine Root Cause(s) Develop Consensus for Continuous Improvement Strategy Determine & Implement Corrective Action Sequence Evaluate Results of Corrective Action Institute Process Redesign Process Redesign Risk Factors Manageable? Develop Process Redesign Flow Modify Process Commit to CI. Monitor Process to Assess On-Going Status Select & Define Next Primary Transactional Problem Establish Performance Improvement Objectives Identify Scopes & Targets Assess Internal & External Process Factors Select New Technologies & Methodologies Develop New Process Test New Process ? Evaluate & Communicate New Process No Go Yes No No Yes Go SL3003Ch10Frame Page 236 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC [...]... Objective: reduce production errors/defects SL3003Ch10Frame Page 237 Friday, November 9, 2001 1:52 PM Identify Internal (Operational) & External (Customer) Production Problems Process Analysis F SL3003Ch10Frame Page 238 Friday, November 9, 2001 1:52 PM 238 The Manufacturing Handbook of Best Practices 10. 4.4 PHASE 1 — MODEL F This model (Figure 10. 7) is best understood by beginning its examination at the... The Manufacturing Handbook of Best Practices Team ID 2 or More Points with Requirement for Knowledge of Cycle (Elapsed) Time Team Creates Data Collection Forms Stakeholders Collect Data Using New Forms Team Reduces Data to Usable Statistics & Statistical Graphs Team Compares Resulting Stats (Baseline) to Stats of Similar Processes in Other Organizations (Benchmarking) Benchmarking G Purpose of Documentation... regarding the quality of the process outputs? How often? © 2002 by CRC Press LLC SL3003Ch10Frame Page 242 Friday, November 9, 2001 1:52 PM 242 The Manufacturing Handbook of Best Practices A.6 DESCRIBE ALL THE INPUTS TO THE WORK ACTIVITY • • • • • • • • A.7 What inputs are sourced from outside or inside the organization? Which inputs are products and which are services? Do any of the inputs have shelf... a current file of process flow charts to increase the likelihood of product consistency The training path is presented to remind a team of the need for current documentation for training new or recently transferred employees The facilitator should encourage the team to identify two or more points within the process where specific knowledge of cycle time (elapsed time) is needed As a rule of thumb, the... their new counterparts so as to develop the new process The new process is tested using all the steps of phase 2 to decide whether it is as good as or better than the original process If the decision is favorable, then it is communicated to all the process stakeholders and we return to the final step of phase 3-A If the decision is a “no go,” then we must return to the first step of phase 2 10. 4.7 PAIN —... examination at the top left and moving to the right or left by following the arrowheads Phase 1 of the model starts with the identification of both internal (operational) and external (customer) problems This can be as simple as developing a comprehensive listing of problems drawn from a specific department, multiple departments (also known as cross-functional), a single division, multiple divisions, or from... points of handoff between process stakeholders With the cycle time points selected, the facilitator should assist the team to create data collection forms, one for each of the selected process points Individual process stakeholders, using the newly created data collection forms, will collect 100 to 200 data values These values should be recorded on the forms © 2002 by CRC Press LLC SL3003Ch10Frame... Developing a consensus strategy for CI follows the identification of the root cause(s) Next, a corrective action sequence is determined and implemented Evaluation of the data generated and collected subsequent to the introduction of the corrective actions should reveal the wisdom of the corrective action sequence © 2002 by CRC Press LLC SL3003Ch10Frame Page 239 Friday, November 9, 2001 1:52 PM Process Analysis... improvement objectives followed by quantification of the target values Phase 3-B continues with the assessment of germane or pertinent internal and external process factors These are the factors that have a high potential of contributing to the success or failure of the process redesign effort At this point the team should turn its attention to the selection of new technologies or methods that may replace... cost, and schedule This completes phase 2 10. 4.6 PHASE 3 Phase 3 begins with a decision, i.e., is the process competitive? If the decision is in the affirmative, then we track phase 3-A in which continuous improvement (Kaizen) of the process is appropriate and should be instituted Continuous improvement (CI) begins with the determination of the root cause(s) of the original problem Developing a consensus . Press LLC 228 The Manufacturing Handbook of Best Practices • Process Map. A two-dimensional version of a process flow chart that also portrays handoffs and receipts of products or services. SL3003Ch10Frame Page 237 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 238 The Manufacturing Handbook of Best Practices 10. 4.4 P HASE 1 — M ODEL F This model (Figure 10. 7). Reduction Controllable Expenses Clean-Up Expenses Miscellaneous Expenses D SL3003Ch10Frame Page 235 Friday, November 9, 2001 1:52 PM © 2002 by CRC Press LLC 236 The Manufacturing Handbook of Best Practices FIGURE 10. 6 Objective: reduce transactional