Manufacturing Quarter Panels - Case Study Example

Decision Theory Harvard Business School Case - The Burred Panels The Background The case refers to a reputed auto company that manufactures quarter panels; and other smaller components for two of their well-known models, Panther and Cheetah. The plant has strength of 3000 employees who work in day or night shifts. The quarter panels are manufactured on four production lines. Each line is headed by a huge hydraulic press. The typical manufacturing set-up is shown in Figure 1. Figure 1: The Assembly Line for Manufacture of Quarter Panels This interesting case study has been presented as a dialog between principal players in the plant. It is therefore pertinent to get to know them and their primary know traits, as listed in Table: 1. Table 1: Key Players in Decision Making Oscar Burger Plant Manager Tough, willingness to listen to others, perceived antiunion by employees Robert Polk Prod. Chief Technically sound, quick tongued, hardnosed driver, perceived antiunion by employees Ben Peters QC Manager Reserved, quiet, cautious when dealing with others, confident with his figures Ralph Coggin IR Manager Sympathetic to employees, typical HR Manager, relies on human relations while dealing with union Joyce Luane Scheduling Supervisor Persistent, analytical and systematic. She has had training in problem analysis procedure, but lacks experience While the players referred above played a major role in problem analysis, there are others, who played a role in building the case, as listed in Table: 2. Table 2: Other Players in Case Study George Adams Supervisor Line #1 Steady, solid, well respected by his team James Farrell Supervisor Line #2 Irascible, ambitious, puritanical, very antiunion Henry Dawson Supervisor Line #3 Patient, warm hearted, genuinely liked by his team Otto Henschel Supervisor Line #4 Aloof, cool, a bit ponderous; neither liked nor disliked by his team Andy Patella Shop Steward Antagonistic to management, and eager to prove his power; has developed rapport with IR Manager Problematic Morning One morning, Production Chief Bob reports that nearly 10% of the panels coming-off production lines, #1 and #2, subsequently even line #4, are being rejected by quality control department for presence of burrs and roughness. An emergency meeting is convened by the Plant Manager, Oscar. The highlights of the discussions are as follows: Strong suspicion of sabotage by the draw-press operators on production lines. This lead gains attention owing to perceived anger over the suspension of a fellow worker Joe Valenti just the previous day, by the hot headed supervisor, James Farrell, who accused Joe of drinking on the job. The effort to rejoin duty by Joe was also thwarted by James. The shop steward Andy Patella, threatens to call a strike if the James is not reprimanded for his arbitrary action, and also if Valenti is not reinstated. There has apparently been no direct communication between James and Coggin, the Industrial Relations Manager regarding Joe’s suspension; and Andy had protested the suspension when he learnt of it. Inspection of incoming steel sheets does not reveal any abnormality. This has been confirmed by line, #1 and #2, supervisors. Maintenance troubleshooters did not find anything wrong with press, hydraulic system or the die. The plant manager collects as much information as possible during the meeting, and then adjourns the meeting until the next morning. In the meanwhile, he hopes he can decide what to do. Plant Manager’s Prognosis Oscar sees two alternatives: Back up James (the supervisor) and risk a strike that might be stopped by injunction. Avoid a strike by undercutting the supervisor, reinstating Valenti, and asking the workers on the line to cooperate in eliminating the excessive rejects. Oscar hopes, he can find another, better alternative, before the second meeting with his managers the following morning. Problem Analysis The follow-up meeting, next morning, kicks up the problem analysis phase. Joyce Luane, who has been invited to attend the meeting, puts the matter in perspective, thanks to some straight objective thinking from her. But before Joyce could intervene into discussions, the plant manager, and his team, exhibited what in problem-solving arena, is called a grave error, that of ‘jumping into conclusions’ about the cause of a problem.. The Grave Error Syndrome Oscar, the plant manager, commenced the meeting by checking on the information he had sought at the previous meeting. These include: Lines #1, #2, and #4 continue to churn rejects exceeding allowable limit of 2%. Line #3 is free of defects. Incidentally, line #3 manufactures panels for Cheetah; whereas, the other production lines are engaged with panels for Panther. There are apparent misgivings that stamping on Cheetah panel is easier than Panther. Engineering department confirms that there is no difference between stamping time for any of the four production lines. Coggin reaffirms the demand of local union agent for an additional 10 minutes of wash-up time. Having satisfied himself, Oscar reinforced his assumption that sabotage was the cause of the high reject rates on the panels. He then made several decisions: which he judged would be capable of taking care of both the reject problem and the labor difficulties. These were: Reprimand James Farrell Reinstate Valenti Jack-up other supervisors so they’ll catch any person trying to produce rejects deliberately Advise team to cooperate and keep rejects to within tolerance Advise Andy not to incite team in the hope of getting a hot issue for new contract negotiations. If the union threatens strike, then the company will get an injunction The Kepner-Tregoe approach (Stryker, 2000; Kepner & Tregoe, 1981) Nipping the fallacy What’s the Problem anyway, asked Joyce? Here are versions of problems stated by key players: Oscar: Unacceptable level of panel rejections bordering with touchy union situation Bob: Discipline in the team, and being too soft with the union Coggin: Train the supervisors in their responsibilities and communication within the team and company (obvious reference to James for not informing him of Valenti’s suspension) Joyce lists the six problems indicated above. Oscar prefers to see these problems as on big problem (another error in problem solving). Joyce further observed: the key players were mixing up problems with decisions. For example: Bob: Lack of discipline (stated as the problem) is the cause for excessive rejects. Coggin: Need for training (stated as the problem) is the cause for low morale. Coggin: Lack of communication (stated as the problem) is the cause for James blunder. Oscar: Assumes all these causes as part of one BIG problem of managing the plant Define the Problem Joyce readily recognizes the confusion stemming in above statements. Technically, a problem is a deviation from some standard norm of desired performance. And, a decision is a choice amongst alternatives. Thus Joyce recognizes: Coggin: is taking about a decision, when he says, “Our real problem is to train supervisors.” Oscar’s: whole problem is not a collection of failures and causes; but a statement of his responsibility for making decisions as head of the plant. So what’s a problem anyways? A problem is something that’s wrong and needs to be fixed. Once the problem is identified, its cause can be discerned. Once the cause is identified, decision making begins. Yes, there could be a series of problems each having its own cause; and, one key problem, that would almost be at the end of a chain of problems and causes. It should also be recognized that the cause of one problem is itself a problem, and its cause, another problem. It’s kind of a stair-stepping sequence. So if the key problem is sorted out, all other problems would disappear. Identify the Key Problem As a first step, the key players identify, panel rejects as the main problem that can have serious repercussions on company’s finance. The next step is to define this problem, precisely. It would help in gathering critical information, and testing any possible causes identified. Define Specifications for the Key Problem Dissect the problem in detail, and get specific facts about it in four different dimensions: what, where, when, and extent. Against each of these dimensions, describe precisely, what the Problem Is; and also, what the Problem Is Not. Table 3 is a snapshot of how the template would look: Table 3: Starting Template to Specify a Problem Deviation Excessive Rejects from Burring on Quarter Panels Is Is Not What deviation object Burrs Panther Panels Any other complaint Cheetah Panels Where (on object observed) Panther Panels Lines #1, #2, #4 Other Parts Line #3 When (on object observed) Line #2: 9:33 AM Line #1: 10:18 AM Line #4: 11:23 AM Any burrs before these times on Lines #1, #2, #4 Line #3 at any time Extent How much How many Bad burring Line #2: 11% Line #1: 17.5% Line #4: 15% Line #3 rejects (within tolerance) In the “Is” column: fill-in those things which are directly affected by the problem. In the “Is Not” column: fill-in those things which are closely related to the problem but not affected by it. Thus, what Joyce has done is to follow a systematic outline to describe precisely both the problem, and what lies outside the problem, but closely related to it. The contrast between ‘Is’ and ‘Is Not’, not only draws a tight boundary around the problem, but strictly limits the amount of information needed for its solution. So we seek only relevant facts. It is pertinent to observe here that: ‘Is’ and ‘Is Not’ focuses on differences, and not similarities. Identify the Cause Use the contrast between ‘Is’ and ‘Is Not’ to identify the cause. Whatever caused the problem produced affects on the ‘Is’ side; therefore, if one thing is affected, and another related thing is not; then, there must be something distinctive, or unique, about the affected thing that sets it apart from the other. The objective is to seek this distinctive factor. Referring to Table 3, we can see a contrast between Panther and Cheetah panels. The Panther panels are affected by the cause, whilst the Cheetah panels are not. So what sets Panther from Cheetah panel? The question is to ask, what is distinctive about those Panther panels? It emerged from Oscar, that shape is a distinctive factor, between Panther and Cheetah. The Panther panels have a deeper draw than Cheetah panels. Table 3 can now be revised with further columns, as shown in Table 4. Table 4: Revision(a) Template to Specify a Problem and its Cause Deviation Excessive Rejects from Burring on Quarter Panels Is Is Not What is Distinctive of the Is? Any Change in this? What deviation object Burrs Panther Panels Any other complaint Cheetah Panels Deep draw -- Where (on object observed) Panther Panels Lines #1, #2, #4 Other Parts Line #3 Deep draw -- When (on object observed) Line #2: 9:33 AM Line #1: 10:18 AM Line #4: 11:23 AM Any burrs before these times on Lines #1, #2, #4 Line #3 at any time Stacks from morning lot blanks used at these times Line#2: 9:30 Line#4: 11:20 Line#1: 10:15 Shift changed between Extent How much How many Bad burring Line #2: 11% Line #1: 17.5% Line #4: 15% Line #3 rejects (within tolerance) Rejection rates not proportional to involvement in Valenti conflict Critical information that Ben brought out was that: Line #1 supervisor had reported bad burrs on his line started just after using up the four stacks of blanks his area already had. The burrs appeared with the blanks loaded with in the morning stacks. The time when the morning stack was loaded onto the Press in different lines were recorded as follows: Line #1: 10:15 AM Line #2: 9:30 AM Line #4: 11:20 AM Line #3: 08:30 AM The Cause of Every Problem is a Change of one Kind or the other This is the heart of the analysis. The distinctions obtained thus far provide areas of possible causes. We must now look for changes that happened owing to these distinctions. The problem analysis team looked at the distinction, “deep draw” and asked: “burred panels must have been caused by some change?” Coming to the dimension, “When”, Joyce pursued with the question, “What changed about those stacks of the morning lot?” A possible answer was that the shift of workers producing the lot of steel blanks changed. This lead was further pursued to test if this plausible cause explains every point in the specifications listed in Table 4, both the ‘Is” and “Is Not” columns across all the dimensions. Considering that something in the morning steel blanks changed, it failed to explain the “what” dimension under “Is” and “Is Not”, since the morning lot steel blanks were used for both Panther and Cheetah panels; and that, Cheetah panels did not produce any burred panels. That means, the change in lots of steel blanks does not fit the bill. Incidentally, in the other dimensions, the team could not identify any convincing change. In problem solving, such dead-ends would often be encountered, indicating that some fresh cause must be searched, or, refinements made in the specs. Re-specifying the Problem Obviously, a change in the workers’ shift between two days is not a material change that could cause burring on panels. There must have been some other noticeable change that must have occurred, and could explain the specifications set-up in Table 4. Deliberations revealed that the change between the lots of steel blanks on two consecutive days was perhaps attributable to receipts of raw material from a new supplier, Zenith. Although, the contract with the new supplier was signed a month earlier, the actual lots began arriving two days back! Discussions also led to the fact that although Zenith supplies met all the technical specifications for raw material, except that, it had a minor change in alloy composition. This would mean that we revise the specifications to state in the ‘when’ dimension, under distinction: Zenith steel blanks began to be used in these times. Revised specification is listed in Table 5. Table 5: Revision(b) Template to Specify a Problem and its Cause Deviation Excessive Rejects from Burring on Quarter Panels Is Is Not What is Distinctive of the Is? Any Change in this? What deviation object Burrs Panther Panels Any other complaint Cheetah Panels Deep draw -- Where (on object observed) Panther Panels Lines #1, #2, #4 Other Parts Line #3 Deep draw -- When (on object observed) Line #2: 9:33 AM Line #1: 10:18 AM Line #4: 11:23 AM Any burrs before these times on Lines #1, #2, #4 Line #3 at any time Zenith Steel blanks began to be used at these times Line#2: 9:30 Line#4: 11:20 Line#1:10:15 New alloy composition in steel blanks Extent How much How many Bad burring Line #2: 11% Line #1: 17.5% Line #4: 15% Line #3 rejects (within tolerance) Rejection rates not proportional to involvement in Valenti conflict Now, how could change in alloy composition, however minor, affect Panther panels, but not, Cheetah panels? Deliberations led to the technical conclusion, which was supported by the Engineering department, that shallow draw would make it easier to use tougher alloy in blanks. And Cheetah panels have indeed shallow draws when compared to Panther. This, the Engineering department explained can cause chances of burrs in deep draw Panther panels! The Engineering Assertion This was verified by the Engineering department of the company. Hypothesis testing was resorted to. Their hypothesis was that the ductility of the new material (owing to alloy changes) was significantly different than that of the old. In order to test their hypothesis, a Brinell hardness test was performed. The purpose of these tests was to determine the hardness and ductility of the materials. The two types of materials were tested. An F-test showed there was no statistical difference in the variance of the old and new material. An equal-variance two-sample t-test revealed that there was a significant difference in the mean hardness of the new and old material. It was therefore concluded that the significant increase in hardness had led to a decrease in ductility of the materials. Also, this different material corresponds unquestioningly with the timing of the jump in defect rate of the deep drawing lines. Thus the less ductile material had led to the higher creation of burrs in the deep drawing processes. Reflection This interesting case study brought out some salient points regarding the problem solving and decision making process in real life. How often, we jump into conclusions based on ‘immediate information’ available to us? How deeply we get clouded by our biases? And sometimes, our problem solving is hindered by our ego block to accept another perspective. How easy it is for us to locate similarities than differences? More significantly, one must learn to remain objective during a problem solving process, especially, when the cause of the problem is not easily detectable. Here is a brief summary of the steps in the Kepner-Tregoe analysis: a) Prepare a decision statement having both an action and a result component b) Establish strategic requirements (Musts), operational objectives (Wants), and restraints (Limits) c) Rank objectives and assign relative weights d) Generate alternatives e) Assign a relative score for each alternative on an objective-by-objective basis f) Calculate weighted score for each alternative and identify top two or three g) List adverse consequences for each top alternative and evaluate probability (high, medium, low) and severity (high, medium, low) h) Make a final, single choice between top alternatives Sherlock Holmes: “It’s quite a three-pipe problem.” Sir Arthur Conan Doyle References Kepner, Charles H., & Tregoe, Benjamin B 1981, New Rational Manager, John Martin Publishing, London. Stryker, Perrin 2000, ‘How to Analyze That Problem’, Harvard Business Review. Read More