ABSTRACT Title of Document: A STRUCTURED METHODOLOGY FOR IDENTIFYING PERFORMANCE METRICS AND MONITORING MAINTENANCE EFFECTIVENESS Maria Mercedes Amoedo, Master?s, 2005 Directed By: Professor Mohammad Modarres Department of Mechanical Engineering Reliability Engineering Program Most current maintenance programs focus on achieving the main goals of maintenance operations: increasing mean time between failures, reducing time to repair and minimizing costs. Some researchers have focused on optimizing these variables. Detailed analyses have been conducted in the fields of equipment wellness, spares administration, planned maintenance and structured organization. Still, many organizations fail to fulfill today?s ambitious objective of guaranteeing operations while achieving high reliability and maintaining safety. A comprehensive method of maintenance assessment that considers key factors and indicators that influence the main goals of maintenance is still sought after. This paper discusses a new approach to performance-based maintenance management. The objective is to determine an integrated reliability management system that provides a method of aligning maintenance operations with the business strategy and monitoring performance of key technical, human and organization goals over time. A STRUCTURED METHODOLOGY FOR IDENTIFYING PERFORMANCE METRICS AND MONITORING MAINTENANCE EFFECTIVENESS By Maria Mercedes Amoedo Thesis submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Master of Science 2005 Advisory Committee: Professor Mohammad Modarres, Chair Professor Ali Mosleh Professor Joseph Bernstein ? Copyright by Maria Mercedes Amoedo 2005 ii Acknowledgements I would like to thank the Ford Argentina Stamping and Body Maintenance staff for providing valuable information to this research. I specially thank Mr. Osvaldo Lerose for his helpful feedback on the trees development. I am also grateful to my parents that always helped me achieve my ambitions and my husband Jan whose unconditional support has always been the foundation to my career and academic success. iii Table of Contents Acknowledgements....................................................................................................... ii Table of Contents.........................................................................................................iii List of Tables ................................................................................................................ v List of Figures.............................................................................................................. vi List of Acronyms ........................................................................................................ vii Chapter 1: Background .................................................................................................1 Chapter 2: Methodology ............................................................................................... 4 2.1 Overview............................................................................................................ 4 2.2 The Balanced Scorecard .................................................................................... 4 2.3 Hierarchical Decomposition Using Goal Tree Analysis.................................... 7 2.4 GTA for the five strategic pillars....................................................................... 8 2.4.1 REACTIVE Pillar ....................................................................................... 8 2.4.2 PROACTIVE Pillar .................................................................................. 12 2.4.3 LOGISTICS Pillar .................................................................................... 17 2.4.4 TRAINING Pillar...................................................................................... 23 2.4.5 PEOPLE Pillar .......................................................................................... 26 2.5 Pillars dependency ........................................................................................... 32 2.6 Metrics Definition............................................................................................ 33 2.7 Weighting Metrics Using the Analytic Hierarchy Process.............................. 35 2.8 Considerations of Feedback............................................................................. 39 2.9 Analysis of scale selection and consistency .................................................... 39 iv Chapter 3: Case Study: Automotive Manufacturing................................................... 43 3.1 Context Definition ........................................................................................... 43 3.2 Sensitivity Analysis ......................................................................................... 46 3.3 Results.............................................................................................................. 51 Chapter 4: Conclusions............................................................................................... 56 Chapter 5: Future Work .............................................................................................. 58 5.1 On the GT decomposition................................................................................ 58 5.2 On the metrics quantification........................................................................... 58 5.3 On the consideration of feedback .................................................................... 58 Appendices.................................................................................................................. 60 Appendix A: Metric Definitions............................................................................. 60 Appendix B: Analytic Hierarchy Process results ................................................... 66 REACTIVE Pillar ............................................................................................... 66 PROACTIVE Pillar ............................................................................................ 72 LOGISTICS Pillar .............................................................................................. 78 TRAINING Pillar................................................................................................ 86 PEOPLE Pillar .................................................................................................... 90 References................................................................................................................... 96 v List of Tables Table 1: Fundamental pillars of the strategy..................................................................6 Table 2: AHP Scale definition.....................................................................................36 Table 3: AHP results for the simplified example on the correct diagnosis attribute...38 Table 4: Automotive industry case study results.........................................................45 Table 5: Observations derived from the attribute sensitivity analysis conducted on the automotive industry case study........................................................................49 vi List of Figures Figure 1: Goal tree hierarchy decomposition.................................................................7 Figure 2: REACTIVE GT............................................................................................10 Figure 3: PROACTIVE GT.........................................................................................13 Figure 4: LOGISTICS GT...........................................................................................18 Figure 5: TRAINING GT............................................................................................24 Figure 6: PEOPLE GT.................................................................................................27 Figure 7: Pillars interdependency................................................................................33 Figure 8: Application of the AHP in metric weighting................................................36 Figure 9: AHP matrices for ?Perform correct diagnosis? simplified example............37 Figure 10: Consistency as a function of the order of the matrix..................................41 Figure 11: Attributes with high contribution to the TRAINING GT goal...................46 Figure 12: Attributes sensitive in the automotive industry case study........................48 Figure 13: Metrics and Pillars dependency for the automotive case study..................52 Figure 14: Relative influence of the resulting metrics over each pillar.......................53 Figure 15: Pillars monitored per metric.......................................................................54 Figure 16: Resulting metrics with their relative weights per Pillar.............................55 vii List of Acronyms AHP: Analytic Hierarchy Process BTS: Built To Schedule CMMS: Computerized Maintenance Management System CSI: Customer Satisfaction Index DFM: Design F or Maintainability FMECA: Failure Mode and Effect Criticality Analysis FTT: First Time Through GT: Goal Tree GTA: Goal Tree Analysis MSI: Maintenance Satisfaction Index MTBF: Mean Time Between Failures MTTR: Mean Time To Repair OEE: Overall Equipment Effectiveness PM: Preventive Maintenance RCM: Reliability Centered Maintenance TPM: Total Productive Maintenance VF: Visual Factory WG: Work Group WO: Work Order 1 Chapter 1: Background Over the past hundred years maintenance management had to rapidly change to keep pace with the increase of complexity in manufacturing processes. In the beginning, equipment maintenance was reduced to optimize the corrective activities in order to minimize downtime. Good performance was dictated by the ability to reduce time to repair. Therefore, the main focus was put on improving human technical skills as well as troubleshooting effectiveness. When reactive maintenance was organized in such a way that failures were immediately found and solved, the need for availability improvement led to preventing failures to occur. The concept of preventive maintenance changes the way of managing maintenance. The objective moves from reactive to proactive maintenance. This means staying ahead of the problem through programmed inspections to find potential failures and eliminate them before they manifest. Different preventive maintenance programs have been implemented. Initially, fixed schedules were developed. These methods did not consider the equipment usage pattern. Consequently, frequent interventions in low utilization equipment represented a waste of resources, while failures still occurred in equipment with higher utilization. In order to develop a customized plan a more careful analysis was needed. This analysis should define the optimum maintenance schedule for each equipment. With customized planning, resources were allocated more efficiently. This led to significant cost reduction and availability improvement. The significant increase in competitive products generated the need to reduce costs and increase quality and reliability. Old techniques were no longer suitable in 2 the new continuous improvement era. One of the initiatives that arose was the Total Productive Maintenance (TPM) [1] . TPM has the objective to prevent failures and quality defects, minimize equipment losses and improve equipment cycle life. The active participation of every part of the organization is the key ingredient for TPM success. Consequently, production personnel participate by conducting inspections and minor interventions on their own equipment. This self-directed maintenance helps detecting equipment malfunctioning in an early stage and provides with important information to maintenance department. Additionally, maintenance force can be assigned to more critical tasks now that minor repairs are handled by production personnel. This innovative approach to maintenance management was a breakthrough. Still, there was a sustained increase in automation and therefore the need for more skilled technicians to ensure equipment performance. Clearly, organizational goals included the reduction of product indirect costs and in most cases hiring was unaffordable so new alternatives in maintenance operations had to be studied. The most recent advances in maintenance management include Design for Maintainability (DFM) [2] and Reliability Centered Maintenance (RCM) [2] . DFM is a proactive approach that aims at reducing the frequency of required repairs, the time to repair and the amount of preventive maintenance interventions. The goal of Design for Maintainability is maintenance prevention. RCM started from the aeronautical industry. Thorough analysis conducted on a group of aircrafts under different maintenance schedules concluded that increasing the frequency of inspection does not necessarily reduce the number of failures. On the 3 contrary, after overhaul the aircrafts would show an increase in the probability of failure due to infant mortality. Additionally, it was found that most failures are related to random events such as poor maintenance practices, overload or improper equipment operation. RCM methodology is based on choosing the most important systems and determining their potential functional failures. With the aid of Failure Mode and Effect Criticality Analysis (FMECA) the most critical causes of failure are identified and an appropriate maintenance plan is developed to control them. This approach admits the ?run to failure? option for those equipment failures that will not represent a significant safety or economical concern on production. The previous discussion shows that maintenance practices evolved to a focus oriented approach where resources are put were they are more needed. Still these initiatives are being implemented among many industries with different levels of success. Evidently, there are other factors making the results widely vary not always properly considered. Success or failure in maintenance management depends on how technical, human and organizational factors are considered. This study will focus on how to integrate these factors and methodically define a set of performance indicators to monitor maintenance operations effectiveness. 4 Chapter 2: Methodology 2.1 Overview The Balanced Scorecard concept [3] will be used to determine the maintenance strategies. This concept will help define the fundamental pillars upon which the overall maintenance operation rests. From these basic pillars, a group of attributes will be derived using a hierarchical decomposition such as the Goal Tree Analysis [4]. Successful implementation and monitoring of these few attributes will lead to more effective management of maintenance operations. A set of metrics must be selected to lead the attributes implementation. These indicators need to monitor the maintenance strategies in such a way that any deviation from the objectives can be detected and immediately corrected. The problem resides in that no attribute can be fully monitored by an isolated metric. As such, a set of indicators would be needed for this purpose. The assignment of each metric to an attribute must be determined through expert judgment. The Analytic Hierarchy Process (AHP) [5] is a powerful tool to formally bring expert judgment to define relevance and importance of each metric to the fulfillment of the attribute. 2.2 The Balanced Scorecard The Balanced Scorecard is a management system that enables the organization to align their vision with the strategy and translate it into action. Its main purpose is to define a set of metrics that will closely monitor the organization performance. The 5 structured methodology allows us to understand the key aspects in maintenance operations preventing the uncontrolled and unfocused selection of performance indicators [6] . In this thesis a model of the Balanced Scorecard has been developed in context of a complex manufacturing plant. The first step in developing the Balanced Scorecard is to define the vision of maintenance operations. This is defined as: Attainment of high performance of people, equipment and processes in maintenance. This ultimate goal is to be accomplished through a methodical strategy that must consider all different aspects of the organization. Therefore, the strategy will be decomposed into fundamental pillars. When selecting the pillars, the first and basic aspect to consider is repairs management. Once a failure occurred the cause must be effectively found and solved. Therefore, the REACTIVE pillar goal must focus on reducing the downtime through minimizing the time to repair. In order to prevent failures to occur in the first place, the focus must change from a reactive to a proactive approach. The PROACTIVE pillar will aim at reducing the amount of failures through appropriate maintenance planning. The goal is to maximize the time between failures. Having good reactive response and effective preventive maintenance (PM) plan is not sufficient without the necessary tools and spares. The LOGISTICS pillar must ensure resource administration including materials, equipment, spares and energy consumption. Therefore this fourth pillar goal is to guarantee resource availability with minimum cost. Even with good planning and having the necessary tools and spares, maintenance personnel must have the appropriate skills to do a 6 quality job. The goal for the TRAINING pillar is to prepare personnel for their job requirements. Finally, it is important to keep in mind that all maintenance related activities are planned, performed or controlled by individuals. Without personnel motivation maintenance results are in jeopardy. PEOPLE pillar is probably the most critical because it is present in all other pillars. Its goal is to increase personnel motivation and performance in order to get the best out of each employee. Table 1 shows the scope of each pillar with its goal definition. Detailed analysis of each pillar will be discussed in the following section. Table 1. Fundamental pillars of the strategy PILLAR SCOPE GOAL REACTIVE Repair action after the failure occurs Minimize time to repair PROACTIVE Planning and monitoring actions to prevent failures Maximize time between failures LOGISTICS Tools, spares and equipment and their availability Guarantee resources availability with minimum cost TRAINING Technical and interpersonal training Prepare personnel for their job requirements PEOPLE Personnel involvement, human performance, safety and workforce planning Get the best performance out of each employee 7 In order to fulfill the overall vision each of the five goals must be realized. It will be considered that each pillar has the same relative importance with respect to the vision accomplishment. 2.3 Hierarchical Decomposition Using Goal Tree Analysis The next step in the balanced scorecard definition is to translate the strategy into action. Goal Tree Analysis (GTA) [4] is the means used in this thesis to perform a hierarchical decomposition of each of the strategic goals. The purpose of the decomposition is to arrive to the lowest measurable function, whereby obtaining the fundamental attributes. In this way, each general goal can be easily managed through the analysis of this few attributes. This simplification is valid given that GTA carefully breaks down the high level goal into subsequent sub goals so that success of all sub goals will guarantee the main goal accomplishment. Figure 1. Goal tree hierarchy decomposition 8 It is important to mention that every subgoal can be eventually decomposed into lower level subgoals. The level of decomposition will be defined intuitively and will mainly depend on the degree to which the attribute can be measured. Therefore, paths that will result from the decomposition may vary in level depth. Figure 1 shows a conceptual diagram of the goal tree (GT). The higher level represented by an oval is the ultimate goal which is decomposed in lower level subgoals until the lowest possible decomposition is met. The shaded blocks represent these fundamental attributes. Note that logical connectors are used to show in which way the combination of various attributes will lead to the goal accomplishment [7] . The AND gate implies that all attributes must be satisfied in order to guarantee the goal success. On the other hand, OR gates indicate that the goal can be met if at least one of the success paths underneath is achieved. Considering this, we must refer to ?alternatives? rather than subgoals given that not all the attributes need to be necessarily met to ensure success at a higher level. The complete decomposition has been conducted considering maintenance operations and management literature and was also based on the authors? judgment. Figures 2 to Figure 6 in the following section show the GTs for each pillar. 2.4 GTA for the five strategic pillars 2.4.1 REACTIVE Pillar Figure 2 shows the hierarchical decomposition for the REACTIVE pillar. There are two possible alternatives to manage a failure depending on the availability 9 of an alternative process. These processes include, backup systems, redundancy, standby equipment and bypass procedures [8] . The OR gate shows that success of either path will lead to the top goal accomplishment. The decision to launch the alternative process will be based on the repair time estimate, the time to switch from normal to alternate operation and the potential loss of production the alternative process represents. Therefore, good communication between maintenance and production personnel is essential to make the best decision. At the same time, clear procedure must be in place to perform a quick change over. An important part of having an effective alternative process in place is its reliability. Stand by and redundant equipment must be in good condition when needed. Even though these installations are rarely used, it is important to have them under planned maintenance. Note that in order to have a good maintenance plan the REACTIVE goal must be satisfied. Figure 2 shows this dependency between REACTIVE and PROACTIVE GTs. The repair path is followed when no alternative process is available or a decision to conduct the repair facing the down time is made. In this case a correct diagnosis followed by an effective repair action is needed. Many variables must work together in order to perform a correct diagnosis. A complete and reliable monitoring system together with appropriate troubleshooting procedures will help detect the failure promptly. Additionally, the technician must have the appropriate knowledge through previous experience or training. As mentioned earlier, good performance also depends on personnel morale and therefore this subgoal will be repeatedly seen throughout the GT. 10 11 Conducting an effective action mainly depends on the technician knowledge and skills. Still, having the ability to conduct the repair is not enough if the proper tools and materials are not available. And even with the skills and resources, the optimum repair action would be carried out if the equipment is easy to maintain. Hard to reach spaces will make the job more difficult, thus increasing the time to repair. Some design approaches such as design for maintainability [2] have this into consideration and provide error proof devices. Having these convenient tools already in the equipment and quick change over procedures can expedite the repair process. The REACTIVE GT shows a detailed decomposition of these many goals in lower level attributes. Some of these attributes deserve a comment. Doc Palmer [8] emphasizes the importance of assigning personnel by skill. Those individuals that are prone to easily find a root cause and promptly implement a solution should be available for production support where time to repair is critical. Generally, these containment actions are highly effective but many lack of quality work given the nature of the repair. On the other hand, preventive and predictive activities should be conducted more carefully. For these interventions, troubleshooting ability is not required but skilled work with high quality finish is essential. Therefore, meticulous technicians should be assigned to planned maintenance tasks. Another important aspect to consider is the work group activities. There are different kinds of meetings that will be explained in more detail when discussing the PEOPLE pillar. The importance of these meetings for the REACTIVE pillar is that they are a source of data analysis and lesson learned communication. 12 2.4.2 PROACTIVE Pillar Figures 3a and 3b show the PROACTIVE pillar decomposition. All planned activities are considered including preventive, predictive and self-directed maintenance, programmed replacements and projects implementation. The latest refers to improvement modifications conducted with maintenance department budget. Preventive, predictive and replacement programs are effective if there is a dynamic schedule, oriented to prevent the loss of the system function. This is the objective of Reliability Centered Maintenance (RCM) [2] . The plan must be routinely evaluated and adjusted based on failure history, condition-based techniques and root cause analysis among other reliability tools. This resource optimization needs appropriate data collection and analysis so having a reliable Computerized Maintenance Management System (CMMS) is essential. Similarly to most management tools, the critical part is not putting the system in place, but maintaining it up to date with all equipment information and analyzing this information routinely. Terry Wireman [9] reinforces the need to have a complete and accurate data in order to support maintenance decisions making process. Therefore, when implementing a CMMS it is important to design an easy to use system, promote personnel involvement and provide necessary resources such as computers and time to enter the data. Optimally, the CMMS will be integrated with other systems in the organization 13 14 15 Planned interventions also need to consider equipment availability. This is especially important when production systems are in continuous operation. In many cases production patterns must be adjusted to support the PM down time. Planning ahead of time production department will ensure an effective intervention without significant production loss. Preventive or predictive activities can be performed by the on site personnel or by a contractor. As discussed in the REACTIVE pillar, the quality of the intervention will depend on the technician skill and morale as well as having the appropriate tools and spares. Additionally, structured procedures must ensure the schedule compliance and also inspection tasks must be clearly defined. This includes not only what to inspect, but also what is considered substandard conditions. This is particularly important for predictive inspections where the variables analyzed increase as the equipments degrade and a threshold value will define the need for replacement. Another important aspect in maintenance inspections is the repair scheduling. For processes that allow short periods of down time it is common to conduct the planned maintenance in two phases. First, the entire equipment is inspected following a detailed checklist. If a substandard condition is found, the technician must decide whether to repair it or program the repair in the near future. This decision will depend on the time needed to conduct the repair, the equipment availability and the tools and spares availability. If the repair is not conducted immediately, the task should be entered in the Work Order (WO) system. This is a very important part of the process that requires discipline. Without proper repair scheduling, the substandard condition 16 can worsen significantly leading to equipment failure before the next PM inspection. The improvement projects include personnel recommendations that can reduce the time to repair, increase the time between failures or reduce the risk of personal injury or property damage. Other projects are derived from manufacturer recommendations or process modifications. For all projects complete engineering specifications must be developed. Additionally, if the project implementation is to be conducted with maintenance workforce, detailed sketches and a list of resources (parts, tools, materials, and manpower) must be prepared. On the other hand, when contractors are involved the implementation plan must be closely followed in order to verify compliance with the specifications. The third type of planned maintenance tasks are the inspections and minor repairs conducted by the operator. This self-directed maintenance approach provides the benefit of discovering equipment problems in an early stage. In order to commit production personnel to add this task to their routine there must be agreement from the union. This step is fundamental when implementing self-directed maintenance. Then, operators must be trained in the inspection checklists as well as in some basic skills to perform minor repairs. An important aspect that is usually overlooked is that conducting daily inspections is time demanding, especially if subsequent interventions are required. Therefore, self-directed activities must be included in production planning. Considering these activities as part of the daily tasks will prevent loss of motivation and operators performance. 17 2.4.3 LOGISTICS Pillar Figures 4a and 4b show the LOGISTICS GT. This pillar focuses mainly on materials, parts and tools availability but it also includes energy as a resource to be administrated. As mentioned earlier, the objective is to ensure resource availability with minimum cost. The latest is the actual challenge. Benjamin W. Niebel [10] defines one of his five primary pillars as ?Cost Reduction? and parts and tools administration is one of many activities to fulfill this goal. In the presented approach, LOGISTICS pillar goal will be accomplished by ensuring equipment, parts and tools availability, optimizing energy consumption and minimizing maintenance inventory. The first condition can be satisfied not only by guaranteeing the part is in stock, but also weather this part is available immediately. Having the part somewhere in a chaotic store will make the repair ineffective and increase the mean time to repair (MTTR). Therefore, great effort should be invested keeping a clear and properly identified storage area. This may include the development of equipment drawings / sketches and a reliable inventory system, as well as applying Visual Factory (VF) procedures. Lack of proper stores administration result in parts unavailability. If storages do not provide with the necessary parts, technicians would start keeping basic spares at hand leading to personal storages generation . Nevertheless, the concept of personal storages should not always be rejected. For large installations where distances are important it would be wise to have materials 18 19 20 near to site. But parts and their quantity must be carefully selected to prevent high inventory. Materials with high circulation and low cost are prone to be in the self- storages. Additionally, spares specified for one particular equipment can also be stored near to site. Even though having parts at hand may significantly reduce the time to repair, multiple storage places may be complex to maintain and control. The lack of organization is a menace for parts administration and it is the main cause of inventory multiplication and high maintenance costs. Therefore, proper analysis of advantages and disadvantages is needed when making the decision to have multiple storage places. In order to ensure the part is in stock when needed, it is essential to conduct adequate planning. Basically, this includes the part list derived from a close analysis conducted in the early design and installation phases. Additionally, one interesting approach that Niebel reinforces is having a parts salvage program [10] . By repairing malfunctioning parts the cost of inventory decreases since a new part is not required. In order to implement a salvage program there must be workshops with the appropriate equipment, sufficient technical skills and clear procedures for repair administration and repair quality assurance. Clearly, the repair would worth the investment if the total repair cost is less than the actual cost of the new part. It is important to notice that the total repair cost not only refers to manpower, parts and materials but there are also hidden costs that usually exceed these tangible values such as opportunity costs. For example, it may take considerable time, skills and resources to repair a failed servomotor from a welding robot. The repair cost can 21 easily surpass the cost of a new servomotor. But if this spare part is not in stock and the arrival time takes weeks, the down time cost generated may be unacceptable given the significant production loss. Planned repairs can represent a great benefit, but it can also increase failure risk considering that the repair does not always leave the part ?as good as new?. Therefore good quality procedures that include testing of the repair of parts must be established. Once the part is certified it can enter the storages and become part of the inventory. But spare parts are one of the three resources that must be administrated. Other important assets that should be controlled are tools and special equipment. Some examples of special equipment include measurement and test equipment, notebooks used for PLC and SLC program access or portable welders. Generally, these types of equipment are expensive and maintenance department own a few. Therefore they deserve special control of their uses and storage. Additionally, they must be under PM schedule. Now that parts, tools and equipment availability was discussed, focus must be on the inventory reduction. It was mentioned that repairing faulty parts helps reduce the amount of parts in stock. Another way to minimize the inventory is by studying parts circulation (for example, how many electrodes are used per week). For this purpose the equipment history must be studied in detail. With this information and the spare acquisition time, a minimum limit is set for that particular part so that when reaching that value a purchase order must be filled. An important cause of high inventory is the multiple types of equipment and 22 vendors. This is common for facilities in expansion where new systems are installed and old equipments are improved. Equipment form vendors that are not certified by the company will certainly have parts list that greatly differs from those that are already specified. Therefore, parts with identical specification but from different suppliers will be duplicated in maintenance stores. Having a list of selected and certified vendors will help minimize this spare parts multiplication reducing the inventory. The final condition for reducing the spare costs is minimizing the need to use them. Well maintained installations will have higher performance and lower failure probability. Consequently the need to replace a defective part will be minimized through proper planned maintenance. As mentioned before, energy consumption will also be treated as a resource to be optimized. This goal will be attained by minimizing energy losses, improving the equipment performance and reducing energy consumption in non operating hours. Some sources of energy loss are water leakage from defective pipe lines, air loss from pneumatic devices, unnecessary power consumption for stand by equipment, etc. One approach for loss control is conducting regular inspections under the preventive maintenance schedule. It is also helpful to have personnel involved in loss detection and reporting. Regarding the non operating hours, a detailed study needs to be conducted to identify the equipment that need to be continuously energized and those that can be powered down. Once the list of equipment to be powered down is defined, clear shut down procedures must be established per equipment. 23 2.4.4 TRAINING Pillar Training is a highly important activity that is usually underestimated. The general believe is that time spent for training is time lost, given that many courses are ineffective and after some weeks the student would probably forget what he was taught. The problem is that this statement is generally true because of the lack of proper planning. Training must be a ?just in time? activity. This means that the person should receive the course when he or she would get the best out of it. Figure 5 shows that training planning should consider the right course for the person. For this purpose, a tool known as training matrix is used. This matrix will relate each employee with the skills and training needed for their job positions. Having defined the matrix, a customized training plan is easily constructed considering not only the courses applicable to the position, but also the adequate level according to the employee?s expertise. Training courses are grouped in four different categories: knowledge base, on the job training, attitudinal and lessons learned. On the job training focuses on skills and tasks directly related to the person?s daily activities. Generally, these courses are taught by more experienced co-workers and are carried out in site. This type of training is especially applicable for new employees or when the person is assigned to a new position. Lessons learned courses are designed to expand individual experiences to the rest of the workforce. The objective is to prevent errors experienced in one application to occur in another one as well as share the best practices among the department. 24 25 Together with the course definition, there must be material preparation and people organization. The course can be prepared within the organization or it can be outsourced. There are advantages and disadvantages in both approaches. Internally designed courses are generally more applicable to the organization needs given that they are customized. But a lot of effort is demanded to prepare the material and installation and usually lack of quality and proper learning methods. On the other hand, external courses are designed by qualified training groups. Additionally, given that many agencies and most manufacturers provide with a set of courses for different customers, they already have the materials and installations ready to use so the course is available immediately. Yet, these courses not always fit the organization particular needs, are less applicable and many times useless. Another disadvantage of external course is that when there are budget cuts, the organization cannot afford contracting external training. Moreover, considering that it is common that the students must attend classes off site overtime is a must which is usually unaffordable in times of recess. These conditions discussed are considered in Figure 5. A similar approach is made when selecting the trainer. Most maintenance management literature reinforces the value of developing interpersonal training. For these cases having internal trainers is the most effective. Proper planning is needed to take the person away from the operation to prepare him as a trainer. For this purpose some maintenance departments have a special team for replacement. These technicians will normally be assigned to improvement tasks such as spare parts repair, equipment testing or projects implementation and will cover the person to be trained up when needed. This same 26 methodology can be seen for ?have people available for training? subgoal. 2.4.5 PEOPLE Pillar As discussed in the previous sections, people morale is a critical aspect that affects most attributes. PEOPLE pillar focuses on getting the best out of every employee. The way to achieve this goal is depicted in Figures 6a and 6b. The first sub goal is to increase employee?s motivation. R.F. Pagano [11] indicates that a person is mainly concerned about self-esteem, independence, self-actualization, and recognition. From this perspective, defining challenging objectives is an important aspect for self-esteem. Additionally, good communication of these objectives as well as departamental and organizational objectives is essential to make the employee understand and become part of the company?s vision. But sharing the goals with the employees will make no difference if there is not an established recognition plan that would reward the individual that actively participates in the results improvement. Independence and self-actualization are two parameters that must be analized when assigning roles and responsabilities. Individuals that are overqualified for their job position will find it difficult to learn something new leading to loss of motivation. But if they are underqulified, they will feel frustrated also leading to motivation problems. In conclusion, the supervisor must ensure that the person is comfortable in his position. Last but not least, there ust be a propert benefits and compensation plan that would be suitable for each employee?s experience and expectations. The program must be aligned with the employee development plan. 27 28 29 Another sub goal in the PEOPLE GT is increase management involvement Note that the term involvement was chosen instead of commitment. Managers can be fully committed to the Organization?s objectives but they need to communicate this commitment to his subordinates in a clear and consistent way. To be consistent, they must give the example by following the standards and procedures established. Also, managers must actively participate in work group meetings. They must understand what the team?s needs are and offer support in order for them to succeed. Additionally, it is important for people to realize that the manager and other supervisors are concerned about day to day activities, so it is important to promote regular visits of managers to the plant. In addition to employees increased motivation, plant touring would allow the managers to get in touch with real problems that are being experienced. The last item in management involvement is to develop highly qualified managers. A proper selection needs to be done from the very beginning, based on the applicant experience and leadership skills. The person to be assigned to this position can be either promoted or hired. Either way, a thorough training program must be provided to enhance technical and personal skills. Employment planning is another activity to consider. It was already mentioned the importance of assigning each employee to the appropriate position. Another important task is to distribute the personnel in order to ensure shift coverage. This is a complex analysis that must balance the need for reactive maintenance technicians in the productive shift with a group of serviceman that will work on 30 pending work orders and the proactive team that will perform preventive and predictive activities. The fact that people will retire some day is commonly neglected. Given that these people are generally highly experienced employees their separation from the organization generates an important knowledge drain. To prevent this situation, an early plan must be developed to prepare new employees for the future vacant position. Based on the previous parameters, the human resources budget should be assigned properly. For example, most organizations have a higher compensation for people that work the night shift. Imagine that the night shift will perform the preventive and predictive tasks. If the strategy is to have 70% of the workforce in proactive activities, therefore this percentage of technicians will be in the night shift. Therefore, enough budget must be assigned to cover the excess of salary compensation for night shift personnel. When discussing motivation it was mentioned the need of proper objectives communication. Communication is a primary subgoal of the PEOPLE pillar decomposition and that is why the ?Communicate objectives effectively? attribute is addressed to ?Induce effective communication? subgoal. Communication must be established bottom up, top down and also laterally. This means that superiors must communicate with their subordinates as well as subordinates need a means to communicate with their superiors. Additionally, communication among co-workers must also be excelled. 31 Work group activities are a good environment to share opinions and discuss problems as well as are suitable to cascade high level objectives. That is why it is essential that these teams are conformed by cross-functional individuals as well as different hierarchies. Three types of meetings are considered. Work group meetings refer to cellular manufacturing teams. These meetings are usually held on a weekly basis. Most participants are from production department with one or two representatives from the supportive areas (maintenance and logistics). People from other areas of interest such as safety, quality or manufacturing are requested to participate if needed. These meeting are always programmed since operations must be stopped in order to gather all the production team. Therefore they have a specific agenda that includes different issues of the area performance (volume, quality, ergonomics, safety, down time, etc.) The second type of meeting is the in site meeting. These are held daily and last only a few minutes. They are conducted in the site while the area is in operation and only a couple of production operators participate together with the maintenance technician and generally the maintenance supervisor and engineer. The main objective is to discuss equipment and installation maintenance issues. Therefore the focus is on reviewing the production log in order to improve the system performance. Communication must be very precise between shifts. Detailed description of the problems faced during the shift of operation must be delivered to the corrective and preventive teams. Similarly, the shift responsible for the PM and system start up must inform any modifications performed in the equipment or anomalies found during setup procedures to the operations shift. 32 The last but probably most important aspect to analyze is the work environment. Safety and health are two conditions that must be guaranteed to every employee. Most organizations have a specific safety department that exclusively focuses on ensuring safe and healthy working conditions. Safety practices are an extensive field of study and will not be explained in detail in this work. For further information please refer to reference [12]. Another important component in a good work environment is resources availability. Both time and tools and spares are considered in Figure 6b. Assigning technicians to different areas is a critical task that must always consider the optimum operator / machine ratio. The analysis must relate the level of complexity of the area with technician skills and familiarity with the equipment. For example, for automatic lines that share electrical and mechanical equipment, at least two technicians must be assigned (one electrical and one mechanical). If it is a complex installation with several equipment, it might be needed to assign more maintenance people, especially if it is a critical system in the process. 2.5 Pillars dependency One important characteristic is that most trees end with fundamental attributes that are common among pillar GTs. The REACTIVE pillar in Figure 2 shows that in order to ensure an effective action, tools and spares must be available. This can only be done through the LOGISTICS pillar. Similarly, personnel competence will be enhanced through proper TRAINING as well as personnel morale will depend on the success of the PEOPLE pillar. This means that the achievement of REACTIVE pillar is directly dependent on LOGISTICS, TRAINING and PEOPLE pillars. This 33 dependency is repeatedly seen in most pillars as represented in Figure 7, showing a feedback process. Note that REACTIVE depends on all other pillars, while TRAINING is completely independent. The evident interdependency among five pillars determines the importance of achieving all the goals simultaneously. This conceptual result reinforces the assumption of assigning equal importance to each pillar. Figure 7. Pillars interdependency 2.6 Metrics Definition Having developed the GTs, the focus shifts to the metric selection. In Figures 2 through 6 all fundamental attributes (shaded boxes) have a set of metrics assigned. There are few cases in which a specific performance indicator can fully monitor a particular attribute. On the contrary, it is more likely that several metrics would be 34 needed to describe a behavior. For example for the ?Develop easy to maintain equipment? attribute in the REACTIVE tree (Figure 2) three metrics are considered: Mean Time To Repair (MTTR), Number of Accidents or incidents, and Maintenance Satisfaction Index (MSI). It is expected that as the equipment becomes easier to maintain, both the time to repair and number of accidents or incidents decrease, while the satisfaction index increases. But there is one question that still remains: In what proportion does each metric represent the attribute fulfillment? This question cannot be answered in a generic way. There are several context dependent situations that vary from application to application. Additionally, even though some aspects of maintenance practices are shared among different industries, there are some characteristics that differ considerably. For example, safety factors are probably the most critical in nuclear industry while reliability without regular inspections is essential for aerospace projects. The GTs resulted in a total of 22 metrics that are seen simultaneously in the five pillars and in different levels of decomposition. Appendix A lists these metrics with their definition. The GTs are developed considering all important aspects of maintenance practices. This general model is later customized to suit particular applications. The customization process will be carried out by assigning relative weights to each metric with respect to the attribute it monitors and also through weighting of the different alternative paths to achieving the pillars. 35 2.7 Weighting Metrics Using the Analytic Hierarchy Process Assigning weight to the metrics is based on expert judgment. It is context dependent and thus depends on the industry for which the trees are being used. When analyzing the context one must understand the economical, social, political and cultural background as well as personnel competence, resource availability and equipment conditions. After considering all these variables, the Analytic Hierarchy Process (AHP) [5] can be used to determine the relative weight of the metrics. The AHP is a decision making process to set priorities and to make the best decision when qualitative aspects of a decision must be considered. It is a systematic method for comparing a list of objectives or alternatives that reduces complex decisions to a series of one-to-one comparisons. The first step of the process is to determine the relative strengths of the metrics in monitoring the attribute. The process consists of conducting a pairwise comparison of the metrics by posing the following question: Is M 1 metric preferred (or more important) over M 2 metric in measuring the attribute? At what level of intensity? The level of intensity can be subjectively assigned through a numeric scale ranging from 1 to 9, where 1 indicates equal importance and 9 absolute importance of M 1 over M 2 . Table 2 shows the scale definition proposed by Saaty [5] . There is also a need to make a comparison among all associated attributes in meeting the higher-level goals. Figure 8 visualizes the comparison procedure. This block diagram shows a simplified example from the REACTIVE pillar. In order to perform a correct diagnosis of a failure, four conditions must be satisfied: increase 36 Table 2. AHP Scale definition Intensity of importance Definition Explanation 1 Equally importance of both elements Two elements contribute equally to the attribute 3 Weak importance of one element over another Experience and judgment slightly favor on element over another 5 Essential or strong importance of one element over another Experience and judgment strongly favor on element over another 7 Demonstrated importance of one element over another An element is strongly favored and its dominance is demonstrated in practice 9 Absolute importance of one element over another The evidence favoring one element over another is of the highest possible order of affirmation personnel morale, improve personnel competence, develop a reliable monitoring system and implement troubleshooting procedures. Each of these attributes can be measured by one or more of the following metrics: Mean Time To Repair (MTTR), Maintenance Satisfaction Index (MSI), Overall Equipment Effectiveness (OEE), and Maintenance Costs. Figure 8. Application of the AHP in metric weighting 37 The methodology consists of evaluating the strength of the metrics in monitoring each of the four attributes with respect to ?Perform correct diagnosis?. One matrix per attribute is constructed as shown in Figure 9. Likewise, a criteria matrix is built to determine the attribute relative importance with respect to the goal. Figure 9. AHP matrices for ?Perform correct diagnosis? simplified example The complete solution will determine the metrics ranking with respect to the ?Perform correct diagnosis? goal as summarized in Table 3. The attributes ranking appears in the first column while the metric ranking is indicated in each corresponding row. The overall ranking will be determined by combining each metric weight with the respective attribute weight. The AHP result for this example 38 shows that MTTR is the most representative metric in measuring ?Perform correct diagnosis? attribute, followed by OEE, then MSI and finally Maintenance Costs. Having these weights assigned, the problem is reduced by one level of decomposition and we move one step upward in the AHP analysis. Table 3. AHP results for the simplified example on the correct diagnosis attribute MTTR MSI OEE Costs 0.076 Morale 0.169 0.615 0.169 0.047 0.261 Personnel 0.564 0.118 0.263 0.055 0.513 Monitoring 0.564 0.118 0.263 0.055 0.150 Troubleshooting 0.564 0.118 0.263 0.055 Final Result 0.534 0.156 0.256 0.054 It is important to mention that there are generally too many metrics involved in the comparison. In order to transfer a limited set of metrics to the upper level, only those with high contribution are selected. The limit is imposed considering the Pareto rule of 80-20. For the analyzed example, the sum of MTTR, OEE and MSI contribution is 94.6 % and therefore ?Costs? metric is not considered further. The procedure is carried out starting from the lowest level attributes. The set of metrics and their ranking derived in this level will serve as the starting point for the next level comparison and this methodology will continue until reaching the pillar goal. Thus, the final indicators will closely reflect the pillar performance. 39 2.8 Considerations of Feedback It was noted that feedback will be present in the GT model due to the dependency of the pillars. Such dependency leads to the existence of fundamental attributes that correspond to the main goal of a number of trees. For these attributes no metric can be effectively selected given that they will most likely differ in each tree. In order to solve this recursive loop problem, a first set of estimated metrics will be considered for the attributes in question. Once the whole process has been conducted in the five pillars, the resulting indicators will now serve as an input for the second round of calculation. The iterative recalculation continues until no variation is observed in any of the five pillars resulting metrics. Another approach to solve the pillars dependency is to use the Analytic Network Process (ANP) [13] . ANP is an enhanced approach to the AHP that supports dependencies and feedback. This theory adds networks to model dependencies among elements under the comparison process. This methodology was not applied in this study. It is left for future studies the application of the ANP and the analysis of how much the results differ from those obtained by the iteration process. 2.9 Analysis of scale selection and consistency There are two aspects in the methodology that deserve detailed analysis: the scale selection and the level of consistency. Both concepts are closely related. The scale proposed by Saaty ranges from 1 to 9 with clear qualitative definitions for the odd values as shown in Table 2. The intermediate values (2, 4, 6 40 and 8) are used when slight distinction is needed. There are several studies that provide different alternatives in the scale selection [14] . Some suggest quadratic and root square scales while others argue that the geometric scale is preferable. But integer scales yield to unevenly dispersed weights and therefore there is the alternative of a balanced scale where the local weights are evenly dispersed over the weight range [0.1, 0.9]. Clearly, the scale selection is highly subjective. For the purpose of this study, the linear 1-9 scale is chosen given that it is an easy way to represent the common verbal statements that the decision maker utilizes when making the metrics comparison. Nevertheless, this scale intransitive behavior must not be overlooked and the consistency results must be analyzed carefully. Consistency is driven mainly by three factors. First, there must be a transitivity consistency. This means that if A is preferred over B and B is preferred over C, therefore A should be preferred over C. If this relation is not sustained, inconsistency will be generated. Nevertheless, there are real life cases where these types of inconsistencies are present. Such is the case of sport teams. It is not uncommon to see that A defeats B, B defeats C and C defeats A. This is a clear example that shows that inconsistency values must be analyzed carefully before assuming that there is judgment error. The second factor affecting consistency is the numerical weights. If A is 3 times preferred over B and B is 3 times preferred over C, then, A should be 3 x 3 = 9 times preferred over C. Any value that does not arithmetically match this result will generate inconsistencies. This condition can represent an important source of 41 uncertainty particularly for qualitative comparisons (the most commonly used in decision making). For this case, it is important to mention that the scale limit of 9 can also compromise the comparison process consistency. For example, if A is 3 times preferred over B and B is 5 times preferred over C, then A should be 3 x 5 = 15 times preferred over C. This value exceeds the upper limit of 9. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0246810 12 14 16 Order of the matrix C o n s i s t e n c y Figure 10. Consistency as a function of the order of the matrix The third important factor is the size of the comparison matrix. The more elements being compared the greater the inconsistency. Figure 10 shows how the size of the matrix affects the consistency [5] . Note that as the number of elements to be compared increase the consistency value decrease. The problem becomes more complex for larger matrices. There is a psychological limit defined by the human 42 capability of managing a certain amount of elements at the same time. The working memory capacity has been experimentally evaluated and it ranges from 5 to 9 items when full attention is deployed [15] . Having all these aspects into consideration, an acceptable level of consistency has to be defined. The consistency ratio (CR) is determined by a consistency index CI and a random index RI through the following expression: CR = CI / RI CI = ( max ? n ) / n ? 1 The random index RI is tabulated and depends on the number of elements in the matrix: n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 RI 0.0 0.0 0.58 0.9 1.12 1.24 1.32 1.41 1.45 1.49 1.51 1.48 1.56 1.57 1.59 T.L Saaty [5] suggests that a consistency ratio CR smaller than 0.1 or 10% is acceptable and for greater values a matrix revision should be made. In this study, all matrices with CR greater than 10% have been carefully reviewed for transitivity and numerical inconsistencies. But given the important amount of high order matrices and significant metrics differences, there are several cases were the value of CR is greater than 10%. These values of inconsistency are acceptable especially in those cases were mainly a rank order is sought after. 43 Chapter 3: Case Study: Automotive Manufacturing A particular case study was selected in order to put the proposed methodology into practice. The study determined the preferred metrics to monitor Maintenance Operations in Automotive Industries under a complex socio-economical environment. 3.1 Context Definition This case study presents some particular characteristics that define the boundary conditions of the analysis. The following list summarizes these conditions: Stamping and Body Plant One shift of production Equipment in poor operating conditions Annual budget cut Annual head count reduction Minimum overtime Limited parts in stock Strong union representation Extreme currency devaluation making spare parts prices exceed the assigned maintenance budget There is a gap of knowledge between technicians and new technologies installed There is no economical aid from the Company Headquarters or from the Government due to global financial difficulties No budget is assigned for training. There is little or no external training 44 High backlog due to poor equipment conditions and high amount of failures Morale: Due to difficult social and economic situation, people in all hierarchy levels are working under great pressure with low motivation The consideration of these conditions will affect the pairwise comparison of the metrics, but will mostly alter the attributes importance. GTs in Figures 2 to 6 include the results from the attributes weight matrices. For example, in Figure 2 it can be seen that ?Ensure tools and spares availability? together with ?Improve personnel competence? are the most important attributes that must be satisfied to guarantee an effective repair. On the contrary, ?Implement Quick Change Over and Error Proofing techniques? is the least significant. Detailed results from the AHP can be found in Appendix B. Note that the previously mentioned iterative process leads to different results depending on the round of iteration. Additionally, the results are listed from higher to lower resulting weights and only those metrics with higher influence are selected. These most representative metrics are highlighted in the resulting tables. The analysis result for this case study is summarized in Table 4. The metrics for PROACTIVE and LOGISTICS were selected using engineering judgment for the first iteration. From this selection the complete process was repeated deriving the set of leading indicators in the ?2nd iteration?. The highlighted metrics are new in the pillar. 45 Table 4. Automotive industry case study results 1st iteration 2nd iteration 3rd iteration BSC Metrics R E A C T I V E MTTR MTBF OEE PM Plan MSI PM audits CSI OEE MTTR MSI Costs PM audits OEE MTTR MSI PM audits Costs PR O A C T I V E MTBF (*) PM plan (*) PM audits (*) OEE (*) CSI (*) OEE MSI PM audits Overtime MTBF Costs CSI OEE MSI PM audits Overtime MTBF Costs CSI L O G I S T I C S Fill rate (*) Inventory (*) MTTR (*) OEE (*) FTT repair (*) Fill rate Costs MTTR Inventory Items repaired OEE Fill rate Costs OEE Inventory Items repaired MTTR T R A I N I N G Overtime MSI Applicability Accidents MTTR Costs Overtime MSI Applicability Accidents MTTR Costs Overtime MSI Applicability Accidents MTTR Costs PE O PL E MSI OEE WG status MTTR Accidents CSI MSI OEE WG status MTTR Accidents CSI MSI OEE WG status MTTR Accidents CSI MTTR PM Audits OEE CSI Costs MSI Overtime Fill Rate Inventory Applicability Accidents WG status MTBF Items repaired (*) Metrics Estimated by Engineering Judgment After three iterations, there were no further changes in the ranking and the final Balanced Scorecard metrics were obtained. 46 3.2 Sensitivity Analysis Given the high number of fundamental attributes estimated and some complexity of comparison matrices, two different approaches have been selected to conduct a sensitivity analysis. One will focus on the importance of metrics weight and the other one on the importance of attributes weight. Figure 11. Attributes with high contribution to the TRAINING GT goal For the metric sensitivity analysis, the attributes weights are kept constant throughout the tree decomposition. With these values, we will identify the most 47 critical paths in each pillar. These can be calculated by multiplying each attribute weight at the different tree levels. The metrics to be evaluated will be those whose attributes weights are larger than 10% contribution to the main goal. To help visualize this condition the path weights for ?Determine right course for the person? from TRAINING Pillar are calculated. Figure 11 summarizes all resulting weights from this path. In addition to the individual attribute weights, each attribute?s contribution to the TRAINING goal is shown. It is expected that attributes in higher levels have a higher contribution to the goal success. Thicker arrows in Figure 12 indicate those attributes that contribute in more than 10% to the main goal and whose metrics will be considered for the sensitivity analysis. This methodology will lead to a limited set of attributes per pillar. Figure 12 lists the resulting attributes with their absolute influence over the goal and the level at which each attribute belongs. This representation shows that higher-level attributes have a greater influence on the pillar goal, which reinforces the conclusion that the metrics comparisons will be more critical as we move to the upper levels. For each of these attributes, the metric sensitivity will be conducted. The procedure consists of varying the metric weight in one level of importance, for example from 3 to 5 in case of increasing relevance or from 9 to 7 for decreasing weight, given that the applied scale uses five absolute measures of importance (1,3,5,7 and 9). These sensible variations may result in metrics rank modification as well as new weight assignments. The resulting observations derived a group of high sensitive metrics and the corresponding AHP matrices were revised. 48 Level 1 Level 2 Level 3 Level 4 0% 20% 40% 60% Alternative Process No alternative process Quick change over Stand by equipment Correct failure diagnosis Corrective action Monitoring system Planned maintenance Projects Self-directed inspections Appropriate PM planning Effective PM execution Projects planning Projects execution On going rescheduling In site PM execution Contractor PM execution Contractor compliance Parts, equipment & tools Low inventory Energy consumption Asset in stock Prompt asset availability New material in stock Adequate assets planning Emergency requisition Asset near to site Adequate parts planning Right course for person Material & Iinstallations Prepaired trainer Applicable to the position Adequate course level Internal skilled trainer External skilled trainer Attitudinal training Lessons learned Motivation Management involvement Working environment Good recognition plan Comfortable in position Safe environment R E A C T I V E P E O P L E L O G S T I C S P R O A C T I V E T R A I N I N G Figure 12. Attributes sensitive in the automotive industry case study 49 The second approach considers constant metric weights and varies the attribute relative values. The criteria immediately below the goal will be subjected to analysis given its dominance in the final result. The objective is to determine how minimum variations in criteria weighting will affect the resulting metrics. All attributes from the first level of decomposition are subjected to individual increased and decreased weights. Observations derived from this sensitivity analysis are listed per attribute within each of the five pillars in Table 5. Table 5. Observations derived from the attribute sensitivity analysis conducted on the automotive industry case study PILLAR CRITERIA INCREASING DECREASING REACTIVE Manage repair actions through alternative process - OEE remains the most relevant metric - Maintenance costs is no more representative of this pillar giving place to Overtime - MTTR decreases 45% moving from the second to the fourth place - OEE remains the most relevant metric - MTTR remains the second indicator but increases its relevance in 45% PROACTIVE Maintain a good planned maintenance program - OEE remains the most relevant metric - One metric less to measure the pillar since CSI is excluded - New PM plan metric is considered while Costs is no longer relevant - MSI becomes the most important metric with a 14% increase followed by Overtime - OEE is reduced in 27% moving to the third place - MTBF is no longer considered giving place to BTS performance indicator - PM audits falls 40% moving to the last place Implement improvement projects - MSI becomes the higher influence metric moving OEE to the second place - CSI is excluded giving place to BTS - OEE remains the most relevant metric - Costs is excluded giving place to BTS Carry out self- directed maintenance - OEE remains the most relevant metric - There is a significant increase in CSI relevance (65%) - Costs is excluded giving place to BTS - There is no change in the first three metrics (OEE, MSI and PM audits) - CSI is no longer considered reducing the total amount of used metrics 50 PILLAR CRITERIA INCREASING DECREASING LOGISTICS Ensure parts, equipment & tools availability - There is no change in the first three metrics (Fill rate, Cost and OEE) - Items repaired is no longer considered reducing the total amount of used metrics from six to five - Costs becomes the most important metric due to a 15% decrease in the Fill Rate metric relevance - MTTR is no longer considered giving place to MSI Minimize inventory - There is no change in the first two metrics (Fill Rate and Costs) - MTTR is no longer considered reducing the total amount of used metrics from six to five - There is no change in the first three metrics (Fill Rate, Costs and OEE) - Inventory drops 67% and is no longer considered giving place to MSI Optimize energy consumption - There is no change in the first three metrics (Fill Rate, Costs and OEE) - Even though there is no significant variation in the metrics results, Items repaired gives place to MSI - There is no change in the metrics ranking but the higher significance of the primary indicators results in one less metric needed (MTTR) TRAINING Determine right courses for the person - Given the close final result weights, a slight variation of the attribute weight easily changes the metrics ranking - Overtime moves from first to fifth place - Costs is no longer considered giving place to Understanding metric - There is no change in the first two metrics (Overtime and MSI) - Accidents drops 36% and is no longer considered giving place to Backload metric - Costs increases 27% moving from the sixth to the third place Have appropriate training material and installations - Overtime remains the most important metric - MSI decreases 15% falling from second to fifth place - Overtime and MSI switch first and second places Have prepared trainers - No significant change in ranking or weights - No significant change in ranking or weights Have people available for training - There is no change in the first two metrics (Overtime and MSI) - It is observed a higher predominance of the first two metrics with respect to the rest of the set - Overtime moves from first to third place after decreasing 13% - Applicability increases 13% moving from third to first place - MSI moves from second to fifth place after decreasing 12% 51 PILLAR CRITERIA INCREASING DECREASING PEOPLE Increase employees motivation - Even though there is no significant variation in the metrics results, CSI is no longer considered giving place to Ideas implementation - The main performance indicators gain more relevance - MSI and OEE remain the first two metrics - WG status decreases 22% falling from the third to the sixth place - Absenteeism is added to the set of metrics Increase management involvement - MSI remains the dominant metric - There is no significant variation in the final weights but some slight changes in ranking appear - No significant variation is perceived Provide Employment planning - No significant variation is perceived - No significant variation is perceived Induce effective communication - Even though there is no significant variation in the metrics results, Accidents is no longer considered giving place to Ideas implementation - No significant variation is perceived Provide with a good work environment - There is no change in the first two metrics (MSI and OEE) - WG status decreases in 20% falling from third to sixth place - Even though there is no significant variation in the metrics results, CSI is no longer considered giving place to Ideas implementation Similar to the metric sensitivity analysis discussed earlier, the result from the attribute sensitivity analysis indicated particular matrices to be carefully reviewed. 3.3 Results The application of the methodology to the automotive industry resulted in a small set of metrics to monitor and effectively manage maintenance operations. These metrics listed in Table 4 resulted from the systematic decompositions of some pillars of effective maintenance along with AHP ranking. 52 Figure 13. Metrics and Pillars dependency for the automotive case study Figure 13 helps visualize the dependencies of the resulting metrics with the pillars. For example, CSI will monitor both PROACTIVE and PEOPLE pillars. Similarly, indicators such as Items Repaired, Inventory and Fill Rate will only monitor LOGISTICS pillar. We can also use Figure 13 to depict which metric will reveal a particular pillar performance. Additionally, Figure 14 details the relative importance of each metric to the pillar goal. Note that in Figure 14 a there is a strong dominance of OEE over the other performance indicators. This means that for this particular case study, the REACTIVE pillar can be primarily monitored by the OEE metric. From a management point of view, focusing on improving OEE will result in an important improvement of the REACTIVE pillar. 53 Figure 14. Relative influence of the resulting metrics over each pillar This same analysis can easily be derived from Figure 15. In this figure we can see the how each metric can monitor different pillars. As an example let us focus on REACTIVE 0% 5% 10% 15% 20% 25% 30% 35% MTTR PM Audits OEE Costs MSI PROACTIVE 0% 5% 10% 15% 20% 25% PM Audits OEE CSI Costs MSI Overtime MTBF LOGISTICS 0% 5% 10% 15% 20% 25% 30% MTTR OEE Costs Fill Rate Inventory Items rep. TRAINING 0% 5% 10% 15% 20% MTTR Costs MSI Overtime Applicability Accidents PEOPLE 0% 5% 10% 15% 20% 25% 30% MTTR OEE CSI MSI Accidents WG status a e dc b 54 Accidents. This metric is shown to monitor both PEOPLE and TRAINING pillars. In order to reduce the number of accidents, the resources should be invested in these two pillars with a slight preference on TRAINING. 0% 20% 40% 60% 80% 100% M T T R P M A u d i t s O E E C S I C o s t s M S I O v e r t i m e F i l l R a t e I n v e n t o r y A p p l i c a b i l i t y A c c i d e n t s W G s t a t u s M T B F I t e m s r e p a i r e d PEOPLE TRAINING LOGISTICS PROACTIVE REACTIVE Figure15. Pillars monitored per metric Similarly, transposing this chart we can obtain the resulting performance indicators that will monitor each pillar as shown in Figure 16. This representation is particularly useful for management decision making since it clearly shows the relative importance each metric has in monitoring a particular pillar. Again, the dominance of OEE as the strongest metric for measuring the REACTIVE pillar can easily be seen. Similarly, this figure also shows that MSI metric is the most relevant metric for measuring PEOPLE pillar. 55 MTTR MTTR MTTR MTTR PM Audits PM Audits OEE OEE OEE OEE CSI CSI Costs Costs Costs Costs MSI MSI MSI MSI Overtime Overtime Fill Rate Accidents WG status MTBF Invent. Applicab. Acc. Items rep. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% REACTIVE PROACTIVE LOGISTICS TRAINING PEOPLE MTTR PM Audits OEE CSI Costs MSI Overtime Fill Rate Inventory Applicability Accidents WG status MTBF Items rep. Figure 16. Resulting metrics with their relative weights per Pillar The proposed method shows what should be monitored to maximize performance of maintenance operations. It also provides the fundamentals to a structured and result-oriented management planning. The sensitivity analysis reinforces the importance of the higher-level attributes in the final results indicating that the metric weighting should be conducted more carefully as we approach the top level. Additionally, those metrics that are sensitive for each critical path are identified so that judgments regarding ranking of the metrics can be modified, if necessary. 56 Chapter 4: Conclusions After developing the GTs for each of the five fundamental pillars a group of attributes was derived. Many of these attributes are replicated in more than one pillar showing that there is a close interdependency among the pillars. Moreover, the presence of closed feedback loops indicates the importance of focusing on all pillars simultaneously in order to ensure optimal maintenance performance. It is observed that most attributes share the same indicators. This means that a variation in a single attribute can modify more than one performance indicators. Also, each metric depends on the success of a number of attributes from different pillars. Therefore, the metrics and attributes show a many-to-many relationship. Decision makers are encountered with this complex model often. The GT decomposition followed by the application of the AHP helped clarify the model dependencies. An important advantage of this methodology is that GT decomposition provided with a general model for maintenance operations. The model can be further customized by applying the AHP to fit particular applications. This study provides a complete and integrated methodology for maintenance related activities. The systematic development of the goal trees allows identification of all the main attributes that should be in place for reliable and safe operations. At the same time, the qualitative hierarchical arrangement of the metrics provides a means of selecting those that will better monitor maintenance performance. A case study was performed and results are consistent with expectation. Most of the resulting metrics are suggested by maintenance management literature. This 57 practical application derived these metrics in a methodic way and at the same time provided with relative weights for each of the five pillars of the strategy. 58 Chapter 5: Future Work 5.1 On the GT decomposition The hierarchical decomposition has been conducted by the author and has been reviewed by another maintenance management expert. Given the complexity of different maintenance practices and the diverse industries and applications, future improvement can be made in the GT development. A team of maintenance experts from different industries can be built in order to conduct a thorough revision of the decomposition and ensure it is applicable to all kind of industries. Additionally, a similar group analysis can be conducted by a cross functional team integrated by experts from the fields of Human Resources, Personnel Development and Training and Logistics and Material Handling. 5.2 On the metrics quantification Given that the AHP quantification is performed by expert judgment it tends to be subjective and dependent on the analyst personal experience. In this work the metric and attribute comparison was conducted only by the author to show the methodology. In order to obtain a more objective result an expert elicitation process should be conducted. Similar to the GT decomposition, this team should involve experts in the maintenance field as well as other cross functional areas. 5.3 On the consideration of feedback In the present work the presence of feedback among pillars was solved using an iterative process. Three iterations have been conducted to obtain the resulting 59 performance indicators. The use of ANP [13] has only been mentioned as an alternative to this problem. Future work should focus on implementing the ANP to address feedback and compare these results with those presented in this study. 60 Appendices Appendix A: Metric Definitions Accidents: Defines the number of accidents and incidents per period of time. It is usually monthly kept and includes the year accumulate. This metric focuses on human injury. Some organizations also include property damage as part of this report. An incident is an event that has the potential to cause damage to personnel. Accidents are usually divided by severity into mayor accidents and minor accidents. Therefore, three values are represented: Applicability: As a training metric, it measures how applicable the course is to the student?s job position. It is usually a qualitative indicator. In order to measure the complete training program the average applicability for all the courses on the period is calculated. Numeric applicability values can be assigned to the qualitative statement to facilitate the graphic representation. Then, the sum of these applicability values over the number of courses in the time period is plotted. Backlog: The backlog can be measured with the work order system by keeping appropriate record of the work order status. This metric measures the amount of work orders (WO) opened and pending for the time period and the ratio of closed WO vs. total WO. In order to see if the backlog is increasing or decreasing, the cumulative values are also represented. 61 BTS: BTS stands for Build To Schedule. It is an operational metric that represent how well the plant produced the correct volume, mix, and sequence according to customer requirements. Cases: ?Case? is the name given to a complete failure analysis that includes the study of why the failure happened (MTBF analysis) and why it took that amount of time to implement the corrective action (MTTR analysis). Additionally, one ?case? includes the containment and definitive corrective actions with the corresponding implementation plan. A case is to be close when the root cause was determined and the definitive corrective actions were implemented. This metric shows the number of opened cases versus the total number of failures occurred as well as the ratio of opened versus closed cases. Costs: There are several ways to measure maintenance costs. Two of the most common are Maintenance cost per unit manufactured and Maintenance cost per total manufacturing costs. It is common to see maintenance costs split in materials costs and manpower costs. The latest also includes over time. CSI: CSI stands for Customer Satisfaction Index. This metric will represent how satisfied Production department is with maintenance performance. In order to gather this information, surveys are commonly used. 62 Fill rate: Fill rate is a material flow performance indicator that represents the ratio of parts provided versus parts requested. FTT repair: FTT stands for First Time Through. This metric is usually used in manufacturing to measure the percentage of good units manufactured in the first round. It is calculated by dividing the number of units minus any defects by the total number of units. This concept can be used for repaired items to measure the quality of the repair. In this case, the metrics will represent the ratio of working repaired units divided by the total units that have been repaired. Ideas Implementation: This metric will measure how many improvement ideas have been proposed by the organization personnel and also the percentage of those that were effectively implemented. Items repaired: This is a simple indication of total items repaired over the total items failed. Inventory: There are to variables that must be considered when measuring inventory. The first one is the total number of items in stock and the second one is the total cost of these items. Generally both values are indicated. The stock inactivity can also be included in this metric computing the inactive stock items divided the total stock items. 63 MSI: Maintenance Satisfaction Index (MSI) represents the moral of maintenance employees. Similar to the Customer Satisfaction Index, this indicator is derived from appropriate surveys. MTBF: Mean Time Between Failure measures the breakdown frequency. It is computed as the considered time period divided by the number of breakdowns. MTTR: Mean Time To Repair as its name clearly indicates, shows the average time to repair. It is computed as the total downtime divided by the number of breakdowns. OEE: The Overall Equipment Effectiveness combines three different measures as follows: OEE = Availability x Performance x Quality Where Availability = Operating Time / Planned Production Time Performance = Ideal Cycle Time / (Operating Time / Total Pieces) Quality = Good Pieces / Total Pieces Overtime: Overtime is the amount of time someone works beyond normal working hours. This metric can be represented by total overtime per period of time or by total overtime cost per period of time. PM audits: In order to improve the preventive maintenance practices, TPM establishes an audit system that consists of randomly select 5% of the total PM work 64 orders and check its compliance and also asses the quality of the work performed. The amount of WO audited with open issues vs. total audited WO is plotted. This performance indicator can help identify training needs as well as the need to rebalance the amount of PM activities. PM plan: This metric represents the compliance of the PM plan through the plotting of total PM work orders closed versus total planed PM work orders. System audits: System audit can be performed on computerized maintenance management system, (CMMS) or the material planning system. This performance indicator is similar to PM audits in that 5% of the items are audited and those with observations are divided by the total of the items audited and the result is then plotted. Troubleshooting: Troubleshooting provides a systematic means of searching for the source of a problem so that it can be solved. All critical equipment must have a comprehensive documentation that is used to guide the analyst thourgh the troubleshooting process. This metric shows how many equipment have troubleshooting documentation over the total plant equipment. Understanding: This training performance indicator measures how effectively the concepts explained in a certain course have been transmitted. In order to measure this 65 level of learning it is necessary that all courses include a brief examination to test the participant?s learning process. The average score of the examination is computed per course and the average of the all the courses results per period of time is represented. WG status or Work Group Status: This indicator only applies to those organizations that have work cells in place. There are many ways to measure how mature the work group is and it is highly related to the type of organization. One example of this technique is using the 10 pillars of Lean Manufacturing and measuring the results in each pillar. The more mature the group is the more advanced Lean manufacturing indicators will be. 66 Appendix B: Analytic Hierarchy Process results REACTIVE Pillar ENSURE PROPER DATA COLLECTION AND ANALYSIS 1 st , 2 nd and 3 rd iteration: 0.067 0.391 0.151 0.391 Final WG meetings In site meetings Database Cases Result OEE 0.088 0.289 0.110 0.145 0.192 Cases 0.027 0.029 0.034 0.371 0.163 MTTR 0.088 0.160 0.110 0.145 0.142 BTS 0.034 0.210 0.015 0.081 0.118 MTBF 0.088 0.058 0.110 0.145 0.102 MSI 0.178 0.100 0.033 0.019 0.064 5% Audits 0.009 0.008 0.373 0.008 0.063 CSI 0.178 0.100 0.011 0.019 0.060 Backload 0.022 0.022 0.182 0.037 0.052 WG 0.288 0.023 0.021 0.029 0.043 INDUCE EFFECTIVE COMMUNICATION 1 st , 2 nd and 3 rd iteration: 0.250 0.750 Final Meetings Training Result Overtime 0.014 0.302 0.230 MSI 0.160 0.224 0.208 Applicability 0.013 0.165 0.127 Accidents 0.111 0.111 0.111 WG 0.309 0.023 0.095 MTTR 0.073 0.083 0.080 Ideas implementation 0.222 0.012 0.064 Costs 0.027 0.049 0.044 MTBF 0.073 0.031 0.041 67 KEEP AN UPDATED LESSON LEARNED SYSTEM 1 st , 2 nd and 3 rd iteration: 0.250 0.750 Final Data analysis Communication Result Overtime 0.010 0.281 0.214 MSI 0.065 0.215 0.177 Applicability 0.010 0.158 0.121 Accidents 0.023 0.115 0.092 OEE 0.280 0.025 0.089 MTTR 0.163 0.056 0.083 WG status 0.036 0.085 0.072 Cases 0.214 0.010 0.061 MTBF 0.085 0.036 0.048 BTS 0.115 0.019 0.043 PERFORM CORRECT DIAGNOSIS 1 st , 2 nd and 3 rd iteration: 0.043 0.232 0.541 0.092 0.092 Final Morale Personnel Monitoring Sys. Lesson Learned Troubleshooting Result MTTR 0.114 0.106 0.323 0.055 0.233 0.230 OEE 0.218 0.157 0.240 0.074 0.182 0.199 MSI 0.285 0.212 0.171 0.214 0.108 0.184 Costs 0.021 0.305 0.071 0.014 0.059 0.117 Troubleshooting 0.012 0.009 0.102 0.018 0.292 0.086 Overtime 0.039 0.074 0.019 0.291 0.012 0.057 Applicability 0.012 0.050 0.018 0.154 0.043 0.040 Accidents 0.084 0.035 0.020 0.112 0.035 0.036 WG 0.157 0.022 0.018 0.038 0.018 0.027 CSI 0.059 0.030 0.018 0.028 0.019 0.024 IMPROVE IN SITE PERSONNEL COMPETENCE 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Training Right position Result MSI 0.221 0.224 0.223 OEE 0.030 0.314 0.172 Overtime 0.316 0.024 0.170 MTTR 0.076 0.160 0.118 Applicability 0.164 0.017 0.091 Accidents 0.110 0.032 0.071 CSI 0.016 0.110 0.063 MTBF 0.028 0.077 0.052 Costs 0.039 0.043 0.041 68 IMPROVE PERSONNEL COMPETENCE 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final In site Contractor Result Costs 0.019 0.428 0.224 OEE 0.246 0.184 0.215 MSI 0.362 0.053 0.207 MTTR 0.102 0.184 0.143 Overtime 0.161 0.115 0.138 Applicability 0.069 0.017 0.043 Accidents 0.042 0.019 0.030 GUARANTEE AN EFFECTIVE ACTION 1 st iteration: 0.494 0.036 0.243 0.113 0.113 Final Tools / spares EP / QCO Personnel Morale Maintainability Result MTTR 0.175 0.300 0.126 0.115 0.311 0.176 OEE 0.126 0.214 0.165 0.214 0.220 0.159 Fill rate 0.303 0.109 0.012 0.011 0.010 0.159 MSI 0.065 0.060 0.217 0.281 0.158 0.137 Inventory 0.218 0.024 0.012 0.011 0.021 0.115 Cost 0.038 0.060 0.275 0.016 0.030 0.093 Accidents 0.030 0.103 0.051 0.087 0.069 0.049 Overtime 0.011 0.009 0.090 0.042 0.120 0.046 CSI 0.021 0.087 0.039 0.063 0.053 0.036 WG status 0.014 0.034 0.012 0.158 0.010 0.030 2 nd iteration: 0.344 0.054 0.344 0.129 0.129 Final Tools / spares EP / QCO Personnel Morale Maintainability Result Cost 0.213 0.060 0.280 0.019 0.061 0.183 MTTR 0.153 0.300 0.135 0.119 0.318 0.172 MSI 0.042 0.060 0.221 0.281 0.179 0.153 OEE 0.056 0.214 0.177 0.215 0.243 0.151 Fill rate 0.289 0.109 0.015 0.014 0.017 0.114 Accidents 0.016 0.103 0.071 0.090 0.095 0.060 Inventory 0.111 0.024 0.015 0.014 0.017 0.049 WG status 0.031 0.034 0.015 0.164 0.017 0.041 CSI 0.009 0.087 0.055 0.069 0.034 0.040 Items repaired 0.080 0.009 0.015 0.015 0.017 0.038 69 3 rd iteration: 0.344 0.054 0.344 0.129 0.129 Final Tools / spares EP / QCO Personnel Morale Maintainability Result Cost 0.212 0.056 0.280 0.022 0.061 0.183 MTTR 0.111 0.305 0.135 0.119 0.318 0.158 OEE 0.064 0.217 0.177 0.214 0.243 0.154 MSI 0.041 0.061 0.221 0.281 0.179 0.153 Fill rate 0.288 0.107 0.015 0.014 0.017 0.114 Inventory 0.149 0.020 0.015 0.014 0.017 0.061 Accidents 0.015 0.096 0.071 0.090 0.095 0.059 WG status 0.029 0.032 0.015 0.164 0.017 0.040 CSI 0.009 0.086 0.055 0.069 0.034 0.040 Items repaired 0.080 0.020 0.015 0.014 0.017 0.038 MANAGE REPAIR ACTIONS WITHOUT ALTERNATIVE PROCESS 1 st iteration: 0.500 0.500 Final Diagnosis Action Result MTTR 0.326 0.330 0.328 OEE 0.232 0.235 0.234 MSI 0.166 0.103 0.134 Fill Rate 0.015 0.161 0.088 Costs 0.111 0.049 0.080 TS 0.079 0.019 0.049 Inventory 0.015 0.074 0.044 Overtime 0.056 0.028 0.042 2 nd iteration: 0.500 0.500 Final Diagnosis Action Result MTTR 0.363 0.239 0.301 Cost 0.109 0.369 0.239 OEE 0.247 0.106 0.177 MSI 0.160 0.162 0.161 Troubleshooting 0.074 0.016 0.045 Fill Rate 0.017 0.067 0.042 Accidents 0.031 0.040 0.036 70 3 rd iteration: 0.500 0.500 Final Diagnosis Action Result MTTR 0.364 0.164 0.264 OEE 0.247 0.242 0.245 Cost 0.114 0.357 0.235 MSI 0.168 0.104 0.136 Fill Rate 0.020 0.071 0.045 Troubleshooting 0.066 0.015 0.040 Inventory 0.020 0.048 0.034 IMPLEMENT QUICK CHANGE OVER 1 st , 2 nd and 3 rd iteration: 0.250 0.750 Final Procedures Communication Result WG status 0.116 0.394 0.324 CSI 0.040 0.203 0.162 MSI 0.022 0.203 0.157 BTS 0.424 0.065 0.155 OEE 0.258 0.085 0.128 Accidents 0.081 0.022 0.037 MTTR 0.059 0.029 0.036 MANAGE REPAIR ACTIONS THROUGH ALTERNATIVE PROCESS 1 st iteration: 0.250 0.750 Final QCO Equipment Result MTBF 0.019 0.347 0.265 PM plan 0.019 0.214 0.165 5% audits 0.019 0.214 0.165 CSI 0.237 0.063 0.106 WG 0.332 0.019 0.098 OEE 0.086 0.086 0.086 MSI 0.169 0.039 0.072 BTS 0.120 0.018 0.044 71 2 nd and 3 rd iteration: 0.250 0.750 Final QCO Equipment Result OEE 0.091 0.290 0.240 PM audits 0.015 0.214 0.164 MSI 0.164 0.158 0.159 Overtime 0.015 0.108 0.085 WG 0.281 0.012 0.080 CSI 0.215 0.028 0.075 PM plan 0.015 0.078 0.062 Costs 0.069 0.055 0.058 BTS 0.119 0.020 0.045 MTBF 0.015 0.038 0.033 REACTIVE PILLAR RESULT 1 st iteration: 0.500 0.500 Final Alt. Proc. No Alt. Proc. Result MTTR 0.025 0.291 0.158 MTBF 0.288 0.019 0.153 OEE 0.059 0.222 0.141 PM plan 0.212 0.012 0.112 MSI 0.038 0.161 0.100 5% audits 0.162 0.012 0.087 CSI 0.107 0.051 0.079 Fill Rate 0.017 0.119 0.068 WG status 0.078 0.026 0.052 Costs 0.015 0.086 0.051 2 nd iteration: 0.500 0.500 Final Alt. Proc. No Alt. Proc. Result OEE 0.330 0.184 0.257 MTTR 0.027 0.332 0.179 MSI 0.161 0.123 0.142 Costs 0.019 0.237 0.128 5% audits 0.235 0.014 0.125 Overtime 0.103 0.036 0.069 CSI 0.052 0.052 0.052 WG status 0.073 0.023 0.048 72 3 rd iteration: 0.500 0.500 Final Alt. Proc. No Alt. Proc. Result OEE 0.330 0.250 0.290 MTTR 0.019 0.330 0.174 MSI 0.161 0.122 0.141 PM audits 0.235 0.013 0.124 Costs 0.028 0.171 0.100 Overtime 0.101 0.053 0.077 WG status 0.074 0.022 0.048 CSI 0.052 0.039 0.046 PROACTIVE Pillar ENSURE PROPER DATA COLLECTION AND ANALYSIS 1 st , 2 nd and 3 rd iteration: 0.067 0.391 0.151 0.391 Final WG meetings In site meetings Database Cases Result OEE 0.088 0.289 0.110 0.145 0.192 Cases 0.027 0.029 0.034 0.371 0.163 MTTR 0.088 0.160 0.110 0.145 0.142 BTS 0.034 0.210 0.015 0.081 0.118 MTBF 0.088 0.058 0.110 0.145 0.102 MSI 0.178 0.100 0.033 0.019 0.064 5% Audits 0.009 0.008 0.373 0.008 0.063 CSI 0.178 0.100 0.011 0.019 0.060 Backload 0.022 0.022 0.182 0.037 0.052 WG 0.288 0.023 0.021 0.029 0.043 CONDUCT THOROUGH RELIABILITY ANALYSIS 1 st , 2 nd and 3 rd iteration: 0.258 0.637 0.105 Final Training Data analysis Resources Result OEE 0.037 0.217 0.216 0.171 MTBF 0.060 0.206 0.036 0.151 Cases 0.011 0.206 0.010 0.135 MSI 0.217 0.064 0.283 0.127 MTTR 0.081 0.133 0.121 0.118 Overtime 0.285 0.012 0.052 0.086 BTS 0.024 0.096 0.026 0.070 Accidents 0.116 0.016 0.087 0.050 Applicability 0.152 0.012 0.010 0.048 WG status 0.015 0.036 0.159 0.044 73 KEEP A RELIABLE WO AND CMMS SYSTEM 1 st , 2 nd and 3 rd iteration: 0.150 0.513 0.076 0.261 Final Personnel User friendly Integration Resource Result MSI 0.284 0.219 0.020 0.197 0.208 WO system Audits 0.015 0.326 0.310 0.032 0.201 Overtime 0.039 0.084 0.145 0.315 0.142 CSI 0.058 0.173 0.020 0.103 0.126 OEE 0.217 0.023 0.053 0.102 0.075 Backload 0.021 0.051 0.086 0.114 0.066 WG status 0.160 0.053 0.092 0.015 0.062 Costs 0.021 0.035 0.212 0.047 0.050 MTTR 0.105 0.023 0.050 0.059 0.047 Accidents 0.080 0.013 0.011 0.016 0.024 CONDUCT APPROPRIATE PLANNING 1 st , 2 nd and 3 rd iteration: 0.066 0.257 0.042 0.104 0.404 0.127 Final WO Reliability Cases Bottle neck Rescheduling Equipment avail. Result OEE 0.088 0.287 0.204 0.194 0.166 0.185 0.199 MBF 0.027 0.219 0.204 0.119 0.206 0.114 0.177 PM plan 0.042 0.101 0.091 0.113 0.289 0.049 0.167 Overtime 0.156 0.027 0.009 0.264 0.020 0.300 0.091 Cases 0.011 0.162 0.284 0.056 0.058 0.011 0.085 BTS 0.012 0.022 0.077 0.123 0.076 0.185 0.077 CSI 0.115 0.011 0.025 0.059 0.095 0.058 0.063 MSI 0.287 0.062 0.037 0.025 0.036 0.068 0.062 Costs 0.042 0.074 0.053 0.036 0.045 0.020 0.048 WO sys. Audits 0.219 0.036 0.015 0.010 0.009 0.011 0.030 IMPROVE IN SITE PERSONNEL COMPETENCE 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Training Right position Result MSI 0.221 0.224 0.223 OEE 0.030 0.314 0.172 Overtime 0.316 0.024 0.170 MTTR 0.076 0.160 0.118 Applicability 0.164 0.017 0.091 Accidents 0.110 0.032 0.071 CSI 0.016 0.110 0.063 MTBF 0.028 0.077 0.052 Costs 0.039 0.043 0.041 74 GUARANTEE IN SITE EFFECTIVE ACTION 1 st , 2 nd and 3 rd iteration: 0.133 0.362 0.133 0.251 0.045 0.077 Final Tools & spares Personnel Morale Compliance Repairs prog. Clear tasks Result MSI 0.072 0.312 0.289 0.040 0.316 0.244 0.204 OEE 0.097 0.238 0.221 0.106 0.155 0.120 0.171 PM audits 0.010 0.053 0.029 0.262 0.017 0.297 0.114 MTTR 0.170 0.171 0.118 0.011 0.017 0.013 0.105 MTBF 0.030 0.083 0.060 0.163 0.054 0.150 0.097 PM plan 0.040 0.028 0.012 0.262 0.017 0.025 0.086 Costs 0.226 0.015 0.053 0.056 0.053 0.052 0.063 WG status 0.053 0.038 0.177 0.018 0.111 0.072 0.059 WO audits 0.010 0.053 0.030 0.070 0.245 0.013 0.054 Fill rate 0.292 0.009 0.011 0.011 0.015 0.013 0.048 ENSURE CONTRACTOR EFFECTIVE ACTION 1 st , 2 nd and 3 rd iteration: 0.750 0.250 Final Compliance Repairs prog. Result PM audits 0.325 0.015 0.248 PM plan 0.221 0.015 0.170 MTBF 0.144 0.029 0.115 WO sys audits 0.073 0.224 0.111 MSI 0.038 0.303 0.104 OEE 0.110 0.083 0.103 CSI 0.058 0.116 0.073 Overtime 0.010 0.166 0.049 Backload 0.021 0.048 0.028 GUARANTEE AN EFFECTIVE ACTION 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final In site contractor Result PM audits 0.162 0.343 0.252 MSI 0.332 0.074 0.203 OEE 0.227 0.051 0.139 PM plan 0.049 0.227 0.138 MTBF 0.069 0.162 0.116 WO system audits 0.018 0.108 0.063 MTTR 0.108 0.013 0.061 Costs 0.034 0.022 0.028 75 MAINTAIN A GOOD PREDICTIVE PROGRAM, MAINTAIN A GOOD PROGRAMMED INSPECTIONS PROGRAM and MAINTAIN A GOOD PROGRAMMED REPLACEMENTS PROGRAM 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Planning Execution Result OEE 0.329 0.161 0.245 PM audits 0.012 0.331 0.172 MTBF 0.235 0.082 0.158 MSI 0.041 0.236 0.139 PM plan 0.160 0.115 0.138 Overtime 0.107 0.044 0.076 Cases 0.076 0.015 0.046 BTS 0.039 0.015 0.027 IMPROVE IN SITE PERSONNEL COMPETENCE 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Training Right position Result MSI 0.221 0.224 0.223 OEE 0.030 0.314 0.172 Overtime 0.316 0.024 0.170 MTTR 0.076 0.160 0.118 Applicability 0.164 0.017 0.091 Accidents 0.110 0.032 0.071 CSI 0.016 0.110 0.063 MTBF 0.028 0.077 0.052 Costs 0.039 0.043 0.041 GUARANTEE AN EFFECTIVE IN SITE PROYECT EXECUTION 1 st , 2 nd and 3 rd iteration: 0.281 0.584 0.135 Final Equipment Personnel Morale Result MSI 0.044 0.289 0.295 0.221 OEE 0.059 0.221 0.225 0.176 MTTR 0.155 0.121 0.129 0.132 Overtime 0.011 0.159 0.069 0.105 Fill rate 0.286 0.013 0.017 0.090 Cost 0.211 0.030 0.045 0.083 Applicability 0.011 0.092 0.017 0.059 WG status 0.029 0.048 0.169 0.059 Inventory 0.113 0.013 0.017 0.042 Items repaired 0.082 0.013 0.017 0.033 76 GUARANTEE AN EFFECTIVE PROYECT EXECUTION 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final In site Contractor Result MSI 0.336 0.143 0.240 Costs 0.051 0.291 0.171 Audits 0.012 0.291 0.151 OEE 0.230 0.051 0.140 Overtime 0.105 0.078 0.091 MTTR 0.164 0.015 0.089 CSI 0.028 0.116 0.072 Fill Rate 0.075 0.015 0.045 DEVELOP COMPLETE ENGINEERING SPECIFICATIONS 1 st , 2 nd and 3 rd iteration: 0.277 0.122 0.122 0.480 Final Manpower task description Sketches Time mgnt Result Overtime 0.438 0.129 0.146 0.453 0.372 Costs 0.287 0.085 0.230 0.261 0.243 MSI 0.127 0.440 0.478 0.042 0.167 Accidents 0.085 0.288 0.081 0.024 0.080 BTS 0.018 0.021 0.018 0.145 0.079 Backload 0.044 0.037 0.047 0.076 0.059 CONDUCT APPROPRIATE PRYECT PLANNING 1 st , 2 nd and 3 rd iteration: 0.200 0.200 0.600 Final Tools and spares Equip. avail. Engineering Result Overtime 0.011 0.317 0.287 0.238 costs 0.210 0.121 0.219 0.197 MSI 0.045 0.071 0.161 0.120 BTS 0.011 0.234 0.089 0.102 Accidents 0.022 0.019 0.123 0.082 OEE 0.058 0.164 0.060 0.081 Fill rate 0.294 0.019 0.015 0.072 MTTR 0.156 0.019 0.015 0.044 Inventory 0.113 0.019 0.015 0.036 Items repaired 0.079 0.019 0.015 0.029 77 IMPLEMENT IMPROVEMENT PROJECTS 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final planning Execution Result MSI 0.162 0.320 0.241 costs 0.227 0.243 0.235 Overtime 0.332 0.079 0.206 Contractor audits 0.012 0.166 0.089 OEE 0.052 0.118 0.085 BTS 0.108 0.017 0.063 Accidents 0.072 0.017 0.045 Fill rate 0.034 0.039 0.037 CARRY OUT SELF-DIRECTED INSPECTIONS 1 st , 2 nd and 3 rd iteration: 0.172 0.100 0.047 0.267 0.414 Final Union Training Tools & spares Morale Time available Result Overtime 0.300 0.265 0.014 0.054 0.302 0.218 CSI 0.217 0.218 0.069 0.261 0.201 0.215 BTS 0.118 0.068 0.025 0.131 0.179 0.137 WG status 0.147 0.032 0.043 0.105 0.102 0.101 MSI 0.044 0.171 0.055 0.210 0.035 0.098 OEE 0.081 0.026 0.093 0.168 0.055 0.089 Costs 0.055 0.061 0.223 0.031 0.086 0.070 Fill Rate 0.012 0.012 0.312 0.013 0.014 0.027 Applicability 0.012 0.137 0.011 0.013 0.014 0.026 Inventory 0.012 0.012 0.156 0.013 0.014 0.020 PROACTIVE PILLAR RESULT 1 st , 2 nd and 3 rd iteration: 0.600 0.200 0.200 Final PM Projects Self-directed Result OEE 0.288 0.084 0.081 0.206 MSI 0.110 0.289 0.106 0.145 PM audits 0.212 0.010 0.012 0.132 Overtime 0.055 0.152 0.284 0.120 MTBF 0.149 0.039 0.039 0.105 Costs 0.021 0.213 0.063 0.068 CSI 0.028 0.032 0.221 0.067 BTS 0.038 0.053 0.168 0.067 PM plan 0.079 0.010 0.012 0.052 Contractor audits 0.021 0.116 0.012 0.038 78 LOGISTICS Pillar CONTROL STORES INTEGRITY 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Security Auditing Result Inventory 0.471 0.270 0.371 Audits 0.049 0.497 0.273 Costs 0.274 0.133 0.203 Fill Rate 0.130 0.065 0.098 MSI 0.076 0.035 0.056 IMPROVE PERSONNEL COMPETENCE 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Training Right position Result MSI 0.221 0.224 0.223 OEE 0.030 0.314 0.172 Overtime 0.316 0.024 0.170 MTTR 0.076 0.160 0.118 Applicability 0.164 0.017 0.091 Accidents 0.110 0.032 0.071 CSI 0.016 0.110 0.063 MTBF 0.028 0.077 0.052 Costs 0.039 0.043 0.041 ADMINISTRATE REPAIRS EFFICIENTLY 1 st , 2 nd and 3 rd iteration: 0.250 0.750 Final Emergency Planned Result Items repaired 0.010 0.316 0.240 Fill rate 0.061 0.217 0.178 Costs 0.041 0.170 0.138 Backload 0.084 0.119 0.110 MTTR 0.339 0.015 0.096 OEE 0.247 0.015 0.073 Inventory 0.029 0.076 0.064 FTT repair 0.053 0.056 0.055 CSI 0.135 0.015 0.045 79 ENSURE QUALITY OF REPAIR 1 st , 2 nd and 3 rd iteration: 0.429 0.143 0.429 Final Test Equipment Procedures Labs Result FTT repair 0.467 0.186 0.301 0.356 Backload 0.147 0.269 0.422 0.282 PM plan 0.258 0.041 0.080 0.151 Fill Rate 0.044 0.404 0.141 0.137 Costs 0.057 0.075 0.034 0.050 OEE 0.027 0.025 0.022 0.025 ADMINISTRATE PART REPAIRS 1 st iteration: 0.195 0.073 0.463 0.073 0.195 Final Personnel Equipment Repair admin. Workshops Quality assurance Result Items repaired 0.061 0.036 0.287 0.329 0.017 0.175 Fill rate 0.028 0.009 0.219 0.101 0.109 0.136 Backload 0.037 0.029 0.116 0.180 0.216 0.119 OEE 0.244 0.182 0.076 0.060 0.065 0.113 Costs 0.010 0.034 0.170 0.036 0.080 0.101 MSI 0.304 0.099 0.032 0.152 0.034 0.099 FTT repair 0.053 0.053 0.048 0.092 0.283 0.098 PM plan 0.010 0.275 0.026 0.008 0.160 0.066 MTBF 0.079 0.275 0.011 0.017 0.027 0.047 Overtime 0.174 0.009 0.015 0.024 0.008 0.045 2 nd and 3 rd iteration: 0.195 0.073 0.463 0.073 0.195 Final Personnel Equipment Repair admin. Workshops Quality assurance Result Items repaired 0.061 0.014 0.287 0.325 0.018 0.173 Fill rate 0.011 0.013 0.219 0.103 0.110 0.133 Backload 0.037 0.027 0.116 0.183 0.218 0.119 OEE 0.243 0.283 0.075 0.063 0.049 0.117 Costs 0.016 0.068 0.170 0.038 0.081 0.105 MSI 0.303 0.161 0.032 0.154 0.028 0.102 FTT repair 0.056 0.013 0.049 0.087 0.286 0.097 Overtime 0.178 0.116 0.016 0.027 0.009 0.054 PM plan 0.014 0.089 0.026 0.010 0.157 0.052 PM audits 0.080 0.216 0.010 0.010 0.045 0.046 80 CONDUCT ADEQUATE PARTS PLANNING 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final New Repaired Result Fill rate 0.310 0.237 0.273 Items Repaired 0.018 0.366 0.192 Costs 0.310 0.064 0.187 OEE 0.180 0.101 0.141 Backload 0.113 0.161 0.137 MSI 0.052 0.044 0.048 FTT repair 0.018 0.028 0.023 CONDUCT ADEQUATE PLANNING 1 st iteration: 0.200 0.200 0.600 Final Tools Machinery Parts Result Fill rate 0.272 0.029 0.291 0.235 Items repaired 0.040 0.059 0.222 0.153 Costs 0.075 0.173 0.170 0.151 MSI 0.123 0.121 0.075 0.094 OEE 0.092 0.027 0.110 0.090 MTTR 0.207 0.018 0.058 0.080 MTBF 0.013 0.229 0.025 0.063 PM 0.013 0.229 0.025 0.063 Accidents 0.153 0.084 0.008 0.052 CSI 0.013 0.032 0.017 0.019 2 nd and 3 rd iteration: 0.200 0.200 0.600 Final Tools Machinery Parts Result Fill rate 0.306 0.040 0.305 0.252 OEE 0.117 0.308 0.113 0.153 Items repaired 0.049 0.031 0.226 0.152 Costs 0.094 0.067 0.177 0.138 MSI 0.162 0.198 0.079 0.119 MTTR 0.227 0.013 0.055 0.081 PM audits 0.015 0.172 0.015 0.046 Overtime 0.015 0.117 0.015 0.035 PM plan 0.015 0.053 0.015 0.023 81 ENSURE ASSETS IN STOCK 1 st iteration: 0.135 0.584 0.281 Final Stores Planning Emergency req. Result Fill Rate 0.120 0.306 0.228 0.259 Costs 0.166 0.163 0.303 0.203 Items Repaired 0.019 0.227 0.034 0.145 MSI 0.089 0.110 0.056 0.092 MTTR 0.034 0.055 0.166 0.083 OEE 0.015 0.074 0.090 0.071 Inventory 0.312 0.023 0.023 0.062 CSI 0.016 0.032 0.090 0.046 Audits 0.231 0.010 0.011 0.040 2 nd and 3 rd iteration: 0.135 0.584 0.281 Final Stores Planning Emergency req. Result Fill Rate 0.120 0.288 0.228 0.249 Costs 0.166 0.120 0.303 0.178 OEE 0.015 0.226 0.090 0.159 Items Repaired 0.019 0.167 0.034 0.110 MTTR 0.034 0.058 0.166 0.085 MSI 0.089 0.081 0.056 0.075 Inventory 0.312 0.019 0.023 0.060 CSI 0.016 0.031 0.090 0.046 Audits 0.231 0.010 0.011 0.040 MINIMIZE HANDLING TIME 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Identification Procedures Result WG status 0.532 0.478 0.505 MTTR 0.236 0.278 0.257 OEE 0.137 0.139 0.138 MSI 0.075 0.081 0.078 CSI 0.021 0.025 0.023 82 MINIMIZE TIME TO OBTAIN THE SPARE 1 st , 2 nd and 3 rd iteration: 0.750 0.250 Final Near to site Quick handling Result MTTR 0.384 0.257 0.352 Costs 0.252 0.036 0.198 OEE 0.160 0.178 0.165 WG status 0.033 0.404 0.126 Inventory 0.111 0.016 0.087 MSI 0.060 0.109 0.072 ENSURE PARTS EQUIPMENT AND TOOLS AVAILABILITY 1 st iteration: 0.750 0.250 Final In stock Prompt Avail. Result Fill rate 0.358 0.027 0.275 Costs 0.243 0.246 0.244 MTTR 0.067 0.363 0.141 Items repaired 0.165 0.015 0.127 MSI 0.099 0.066 0.091 OEE 0.046 0.174 0.078 WG status 0.022 0.110 0.044 2 nd and 3 rd iteration: 0.750 0.250 Final In stock Prompt Avail. Result Fill rate 0.358 0.027 0.275 Costs 0.243 0.246 0.244 OEE 0.157 0.174 0.161 MTTR 0.071 0.363 0.144 Items repaired 0.104 0.015 0.082 MSI 0.046 0.066 0.051 WG status 0.021 0.110 0.043 83 MINIMIZE NEW MATERIAL IN STOCK 1 st , 2 nd and 3 rd iteration: 0.500 0.500 Final Repairs Reduce quantity Result Fill rate 0.222 0.162 0.192 Inventory 0.023 0.346 0.185 Items repaired 0.311 0.029 0.170 Costs 0.075 0.256 0.166 Backload 0.164 0.066 0.115 OEE 0.107 0.018 0.063 MSI 0.048 0.042 0.045 WG status 0.016 0.052 0.034 FTT 0.034 0.029 0.032 MINIMIZE INVENTORY 1 st iteration: 0.261 0.513 0.076 0.150 Final Unification Minimum new material Reliable record Minimum use Result Fill rate 0.143 0.290 0.068 0.017 0.194 Inventory 0.242 0.214 0.227 0.017 0.193 Costs 0.331 0.127 0.162 0.047 0.171 Items repaired 0.050 0.166 0.040 0.017 0.104 WG status 0.106 0.074 0.099 0.061 0.082 Stock audits 0.053 0.048 0.323 0.015 0.066 PM plan 0.015 0.012 0.013 0.314 0.058 OEE 0.030 0.044 0.044 0.097 0.048 MTBF 0.015 0.012 0.013 0.208 0.042 PM audits 0.015 0.012 0.013 0.208 0.042 2 nd and 3 rd iteration: 0.261 0.513 0.076 0.150 Final Unification Minimum new material Reliable record Minimum use Result Inventory 0.245 0.211 0.230 0.018 0.192 Fill rate 0.142 0.285 0.061 0.018 0.191 Items repaired 0.046 0.161 0.034 0.018 0.100 Costs 0.335 0.122 0.164 0.087 0.176 WG status 0.104 0.056 0.100 0.036 0.069 MSI 0.038 0.020 0.026 0.190 0.051 Backload 0.021 0.078 0.015 0.069 0.057 PM audits 0.011 0.010 0.010 0.243 0.045 Stock audits 0.033 0.024 0.327 0.018 0.048 OEE 0.024 0.034 0.033 0.302 0.072 84 PROMOTE PERSONNEL INVOLVEMENT 1 st, 2 nd and 3 rd iteration: 0.750 0.250 Final Morale Awareness Result MSI 0.309 0.224 0.288 OEE 0.222 0.033 0.175 WG status 0.159 0.023 0.125 Over time 0.036 0.312 0.105 MTTR 0.111 0.073 0.101 Accidents 0.079 0.108 0.086 Applicability 0.012 0.160 0.049 CSI 0.054 0.016 0.044 Costs 0.019 0.051 0.027 MINIMIZE CONSUMPTION IN NON OPERATING HOURS 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final Shut down Personnel involvement Result Costs 0.471 0.023 0.247 MSI 0.102 0.360 0.231 WG status 0.264 0.158 0.211 OEE 0.033 0.245 0.139 Overtime 0.033 0.100 0.066 Accidents 0.066 0.046 0.056 MTTR 0.033 0.068 0.050 MINIMIZE LOSSES 1 st iteration: 0.250 0.750 Final Personnel detection PM program Result PM plan 0.014 0.307 0.233 OEE 0.215 0.125 0.148 MTBF 0.028 0.177 0.140 5% Audits 0.009 0.171 0.131 MSI 0.282 0.037 0.098 CSI 0.044 0.079 0.070 WG status 0.159 0.033 0.065 Overtime 0.111 0.037 0.056 MTTR 0.085 0.014 0.032 Accidents 0.052 0.020 0.028 85 2 nd and 3 rd iteration: 0.250 0.750 Final Detection Control program Result OEE 0.217 0.291 0.273 MSI 0.284 0.160 0.191 PM audits 0.013 0.209 0.160 Overtime 0.113 0.108 0.109 PM plan 0.013 0.076 0.061 WG status 0.165 0.019 0.056 Cost 0.029 0.058 0.051 MTBF 0.020 0.043 0.037 Accidents 0.060 0.024 0.033 MTTR 0.086 0.011 0.029 OPTIMIZE ENERGY CONSUMPTION 1 st iteration: 0.429 0.143 0.429 Final Losses Equipment performance Shut Down Result PM plan 0.332 0.337 0.024 0.201 Costs 0.048 0.041 0.343 0.173 OEE 0.227 0.112 0.132 0.170 MSI 0.068 0.031 0.244 0.138 MTBF 0.162 0.225 0.024 0.112 WG status 0.025 0.012 0.186 0.092 5% Audits 0.101 0.157 0.024 0.076 CSI 0.037 0.085 0.024 0.038 2 nd and 3 rd iteration: 0.429 0.143 0.429 Final Losses Equipment performance Shut Down Result OEE 0.361 0.359 0.113 0.254 MSI 0.245 0.165 0.245 0.234 Costs 0.025 0.047 0.361 0.172 PM audits 0.159 0.244 0.019 0.111 WG status 0.068 0.016 0.167 0.103 Overtime 0.100 0.101 0.077 0.090 PM plan 0.043 0.068 0.019 0.036 86 LOGISTICS PILLAR RESULT 1 st iteration: 0.637 0.258 0.105 Final Available parts Inventory Energy consumption Result Fill rate 0.287 0.156 0.019 0.225 Costs 0.212 0.211 0.228 0.213 MTTR 0.156 0.011 0.019 0.104 Inventory 0.040 0.287 0.019 0.102 Items repaired 0.109 0.113 0.019 0.101 OEE 0.057 0.032 0.183 0.064 MSI 0.079 0.011 0.086 0.062 WG status 0.030 0.082 0.096 0.050 PM plan 0.010 0.044 0.299 0.049 Stock audit 0.020 0.053 0.030 0.029 2 nd and 3 rd iteration: 0.637 0.258 0.105 Final Available parts Inventory Energy consumption Result Fill rate 0.311 0.164 0.019 0.243 Costs 0.214 0.222 0.168 0.211 OEE 0.158 0.061 0.318 0.150 Inventory 0.030 0.311 0.019 0.102 Items repaired 0.073 0.111 0.019 0.077 MTTR 0.111 0.013 0.019 0.076 MSI 0.054 0.033 0.227 0.067 WG status 0.039 0.056 0.092 0.049 PM audit 0.010 0.031 0.120 0.027 TRAINING Pillar DEFINE ON THE JOB TRAINING COURSES 1 st, 2 nd and 3 rd iteration: 0.600 0.200 0.200 Final SKILLS PROACTIVE REACTIVE Result FTT 0.419 0.056 0.096 0.282 Accidents 0.290 0.110 0.080 0.212 MTBF 0.104 0.503 0.025 0.168 MTTR 0.104 0.027 0.453 0.158 CSI 0.029 0.235 0.230 0.110 MSI 0.055 0.070 0.115 0.070 87 DEFINE COURSES APPLICABLE TO THE POSITION 1 st, 2 nd and 3 rd iteration: 0.391 0.391 0.151 0.067 Final Attitudinal Lesson Learned On the job General Result Accidents 0.207 0.194 0.301 0.029 0.204 MTTR 0.099 0.204 0.162 0.089 0.149 MSI 0.252 0.019 0.097 0.213 0.135 TS 0.024 0.278 0.023 0.028 0.123 5%Audits 0.159 0.024 0.043 0.016 0.079 CASES 0.018 0.131 0.024 0.244 0.078 MTBF 0.037 0.052 0.202 0.157 0.076 WG 0.114 0.050 0.015 0.090 0.072 CSI 0.077 0.028 0.121 0.028 0.061 COSTS 0.013 0.020 0.011 0.105 0.022 DETERMINE RIGHT COURSE FOR THE PERSON 1 st, 2 nd and 3 rd iteration: 0.750 0.250 Final Applicable Adequate level Result Accidents 0.298 0.019 0.228 MTTR 0.243 0.130 0.215 MSI 0.168 0.183 0.172 Understanding 0.025 0.458 0.134 TS 0.112 0.031 0.092 5% Audits 0.075 0.039 0.066 MTBF 0.050 0.089 0.060 CASES 0.028 0.051 0.034 ENSURE MATERIALS AND INSTALLATIONS APPLICABILITY 1 st, 2 nd and 3 rd iteration: 0.750 0.250 Final Similar Inst. Course adapt. Result Applicability 0.594 0.501 0.571 Understanding 0.121 0.246 0.152 MTTR 0.104 0.104 0.104 MTBF 0.104 0.104 0.104 MSI 0.078 0.045 0.070 88 USE EXTERNAL MATERIAL AND INSTALLATIONS 1 st, 2 nd and 3 rd iteration: 0.750 0.250 Final Participation Applicable Result Overtime 0.290 0.014 0.221 MTTR 0.165 0.167 0.166 Backload 0.178 0.014 0.137 PM plan 0.152 0.020 0.119 Applicability 0.030 0.305 0.099 Costs 0.101 0.038 0.085 Understanding 0.013 0.226 0.066 MSI 0.045 0.092 0.056 MTBF 0.025 0.124 0.050 DEVELOP MATERIAL AND INSTALLATIONS INTERNALLY 1 st, 2 nd and 3 rd iteration: 0.637 0.258 0.105 Final Equipment Material Logistics Result Costs 0.481 0.076 0.554 0.384 Applicability 0.310 0.440 0.038 0.315 Understanding 0.116 0.265 0.038 0.146 Overtime 0.034 0.154 0.291 0.092 MSI 0.059 0.064 0.080 0.063 HAVE APPROPRIATE TRAINING MATERIAL AND INSTALLATIONS 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final In site Off site Result Overtime 0.118 0.335 0.226 Costs 0.350 0.045 0.197 Applicability 0.255 0.066 0.161 MTTR 0.048 0.245 0.146 Understanding 0.186 0.031 0.108 Backload 0.025 0.166 0.096 PM Plan 0.017 0.113 0.065 89 DEVELOP INTERNAL TRAINER WITHOUT OVERTIME 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final Replacement No replacement Result MSI 0.266 0.165 0.216 Backload 0.339 0.022 0.180 MTTR 0.028 0.298 0.163 OEE 0.028 0.298 0.163 PM plan 0.181 0.022 0.101 CSI 0.061 0.136 0.098 Understanding 0.096 0.061 0.079 HAVE PREPARED TRAINERS 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final External Internal Result Costs 0.362 0.130 0.246 Overtime 0.032 0.376 0.204 Applicability 0.226 0.090 0.158 MSI 0.091 0.221 0.156 Understanding 0.226 0.066 0.146 Backload 0.032 0.080 0.056 MTTR 0.032 0.037 0.034 HAVE PEOPLE AVAILABLE FOR TRAINING 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final overtime No overtime Result Over time 0.445 0.022 0.233 MSI 0.131 0.323 0.227 Backload 0.032 0.230 0.131 Costs 0.226 0.017 0.121 MTTR 0.032 0.164 0.098 OEE 0.032 0.110 0.071 CSI 0.072 0.056 0.064 PM plan 0.032 0.078 0.055 90 TRAINING PILLAR RESULT 1 st, 2 nd and 3 rd iteration: 0.500 0.147 0.288 0.066 Final Right course Installations People availability Trainer Result Overtime 0.014 0.306 0.313 0.220 0.157 MSI 0.126 0.037 0.225 0.114 0.141 Applicability 0.194 0.159 0.028 0.163 0.139 Accidents 0.246 0.014 0.032 0.010 0.135 MTTR 0.170 0.107 0.089 0.042 0.129 Cost 0.037 0.220 0.120 0.297 0.105 Understanding 0.127 0.082 0.014 0.080 0.085 Backload 0.016 0.055 0.167 0.057 0.068 Troubleshooting 0.069 0.019 0.013 0.017 0.042 PEOPLE Pillar INDUCE EFFECTIVE BOTTOM UP COMMUNICATION 1 st, 2 nd and 3 rd iteration: 0.333 0.333 0.333 Final Survey Meetings Personnel suggestions Result Ideas Implementation 0.274 0.266 0.448 0.329 WG status 0.160 0.381 0.149 0.230 MSI 0.376 0.104 0.193 0.224 CSI 0.034 0.104 0.020 0.053 Cost 0.016 0.066 0.073 0.052 MTTR 0.059 0.034 0.050 0.048 MTBF 0.059 0.020 0.027 0.035 Accidents 0.022 0.026 0.041 0.029 INDUCE EFFECTIVE TOP DOWN COMMUNICATION 1 st, 2 nd and 3 rd iteration: 0.637 0.258 0.105 Final Leadership WG participation Cascades Result MSI 0.334 0.206 0.325 0.300 OEE 0.156 0.064 0.146 0.131 MTTR 0.156 0.064 0.146 0.131 MTBF 0.156 0.064 0.146 0.131 CSI 0.084 0.206 0.043 0.112 WG status 0.023 0.332 0.031 0.103 Cost 0.060 0.032 0.146 0.062 Ideas Impl. 0.032 0.032 0.018 0.030 91 INDUCE EFFECTIVE COMMUNICATION AMONG MAINTENANCE 1 st, 2 nd and 3 rd iteration: 0.258 0.637 0.105 Final Between shifts WG meetings Union Result WG status 0.044 0.328 0.073 0.228 MSI 0.292 0.170 0.289 0.214 Ideas 0.014 0.201 0.073 0.139 Accidents 0.049 0.088 0.222 0.092 Backload 0.188 0.033 0.039 0.074 MTTR 0.104 0.056 0.043 0.067 CSI 0.156 0.023 0.039 0.059 PM plan 0.110 0.033 0.016 0.051 MTBF 0.031 0.056 0.016 0.046 Overtime 0.011 0.011 0.191 0.030 INDUCE EFFECTIVE COMMUNICATION WITH OTHERS 1 st, 2 nd and 3 rd iteration: 0.250 0.750 Final WG meetings In site meetings Result OEE 0.076 0.252 0.208 BTS 0.018 0.252 0.194 WG status 0.392 0.019 0.112 CSI 0.146 0.082 0.098 MSI 0.146 0.082 0.098 MTTR 0.076 0.098 0.092 MTBF 0.076 0.072 0.073 Ideas implementation 0.025 0.082 0.068 Accidents 0.044 0.061 0.057 INDUCE EFFECTIVE LATERAL COMMUNICATION 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final Maintenance Others Result WG status 0.307 0.154 0.230 OEE 0.026 0.315 0.171 MSI 0.219 0.088 0.153 BTS 0.010 0.238 0.124 Ideas implementation 0.162 0.030 0.096 Accidents 0.109 0.022 0.066 CSI 0.038 0.088 0.063 MTTR 0.053 0.050 0.052 Backload 0.076 0.014 0.045 92 INDUCE EFFECTIVE COMMUNICATION 1 st, 2 nd and 3 rd iteration: 0.333 0.333 0.333 Final Bottom up Top Down Lateral Result MSI 0.171 0.293 0.152 0.205 WG 0.213 0.056 0.284 0.184 OEE 0.027 0.205 0.217 0.150 Ideas 0.288 0.039 0.079 0.135 MTTR 0.070 0.155 0.019 0.082 CSI 0.117 0.083 0.039 0.080 MTBF 0.052 0.112 0.019 0.061 BTS 0.009 0.011 0.108 0.043 Accidents 0.039 0.029 0.054 0.041 Backload 0.015 0.016 0.028 0.020 COMMUNICATE CLEAR OBJECTIVES 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final Defined objectives Communication Result MSI 0.447 0.371 0.409 WG status 0.026 0.240 0.133 OEE 0.170 0.067 0.118 Cost 0.170 0.024 0.097 MTTR 0.086 0.106 0.096 Ideas Implementation 0.014 0.155 0.085 MTBF 0.086 0.037 0.062 MAKE PERSONNEL BE COMFORTABLE IN THEIR POSITION 1 st, 2 nd and 3 rd iteration: 0.250 0.750 Final training Right position Result OEE 0.044 0.282 0.223 MSI 0.211 0.217 0.216 MTTR 0.080 0.166 0.144 CSI 0.026 0.118 0.095 Overtime 0.293 0.026 0.093 MTBF 0.026 0.087 0.072 Accidents 0.109 0.031 0.051 Applicability 0.161 0.010 0.048 PM plan 0.016 0.042 0.035 Cost 0.034 0.021 0.024 93 INCREASE EMPLOYEES MOTIVATION 1 st, 2 nd and 3 rd iteration: 0.166 0.046 0.443 0.258 0.087 Final Clear objectives Challenging objectives Recognition plan Right position Compensation Result MSI 0.282 0.288 0.285 0.225 0.315 0.272 OEE 0.164 0.071 0.093 0.288 0.052 0.151 WG status 0.215 0.137 0.158 0.015 0.028 0.118 Ideas implementation 0.063 0.213 0.197 0.025 0.027 0.116 MTTR 0.086 0.071 0.093 0.163 0.056 0.106 Costs 0.112 0.033 0.093 0.031 0.221 0.089 Overtime 0.017 0.020 0.014 0.109 0.108 0.048 5% audits 0.033 0.038 0.040 0.046 0.018 0.038 Absenteeism 0.010 0.120 0.017 0.012 0.166 0.032 CSI 0.018 0.009 0.009 0.086 0.008 0.030 CONDUCT APPROPRIATE MANAGER SELECTION 1 st, 2 nd and 3 rd iteration: 0.500 0.500 Final Internal External Result OEE 0.250 0.165 0.208 MTTR 0.250 0.165 0.208 MTBF 0.250 0.165 0.208 cost 0.030 0.374 0.202 MSI 0.133 0.079 0.106 CSI 0.088 0.051 0.069 DEVELOP HIGHLY QUALIFIED MANAGERS 1 st, 2 nd and 3 rd iteration: 0.750 0.250 Final Good selection Training Result MTTR 0.240 0.067 0.197 OEE 0.240 0.028 0.187 MTBF 0.240 0.028 0.187 MSI 0.078 0.240 0.119 Cost 0.130 0.028 0.105 Overtime 0.018 0.337 0.098 Accidents 0.039 0.107 0.056 Applicability 0.013 0.164 0.051 94 SHOW INVOLVEMENT BY GIVING THE EXAMPLE (FOLLOWING STANDARDS) 1 st, 2 nd and 3 rd iteration: 0.750 0.250 Final High morale Awareness Result MSI 0.381 0.229 0.343 Absenteeism 0.225 0.014 0.172 Overtime 0.071 0.331 0.136 Accidents 0.100 0.113 0.104 CSI 0.107 0.023 0.086 MTTR 0.071 0.078 0.073 Applicability 0.013 0.168 0.052 Cost 0.032 0.043 0.035 INCREASE MANAGEMENT INVOLVEMENT 1 st, 2 nd and 3 rd iteration: 0.510 0.072 0.260 0.115 0.043 Final Leadership Walk the plant Meeting participation Give example Prepared managers Result MSI 0.261 0.344 0.173 0.314 0.118 0.244 MTTR 0.164 0.162 0.087 0.052 0.286 0.136 MTBF 0.164 0.162 0.087 0.052 0.181 0.132 OEE 0.164 0.121 0.087 0.052 0.181 0.129 CSI 0.090 0.069 0.173 0.216 0.036 0.122 WG 0.047 0.052 0.297 0.033 0.024 0.110 Cost 0.057 0.016 0.047 0.026 0.091 0.049 Absenteeism 0.024 0.023 0.028 0.174 0.008 0.042 Overtime 0.011 0.009 0.011 0.060 0.059 0.018 5% audits 0.018 0.042 0.011 0.021 0.015 0.018 PROVIDE EMPLOYMENT PLANNING 1 st, 2 nd and 3 rd iteration: 0.076 0.513 0.150 0.261 Final Retired employees Right position Budget Shift coverage Result OEE 0.357 0.305 0.182 0.244 0.275 MTTR 0.174 0.155 0.059 0.122 0.133 MSI 0.041 0.200 0.018 0.036 0.118 CSI 0.078 0.095 0.028 0.147 0.097 Cost 0.126 0.033 0.300 0.015 0.076 PM plan 0.032 0.027 0.071 0.157 0.068 Backload 0.063 0.053 0.027 0.120 0.067 MTBF 0.017 0.077 0.059 0.065 0.067 Overtime 0.011 0.013 0.241 0.075 0.063 FTT repair 0.100 0.042 0.014 0.019 0.036 95 PROVIDE WITH A GOOD WORK ENVIRONMENT 1 st iteration: 0.288 0.147 0.500 0.066 Final O / M ratio Healthy Safe Tools & equip. Result Accidents 0.045 0.119 0.310 0.069 0.190 MSI 0.201 0.271 0.140 0.095 0.174 CSI 0.066 0.271 0.140 0.024 0.130 Absenteeism 0.021 0.119 0.205 0.013 0.127 Overtime 0.287 0.036 0.035 0.009 0.106 MTTR 0.153 0.089 0.059 0.138 0.096 OEE 0.080 0.050 0.073 0.057 0.070 Backload 0.124 0.025 0.019 0.032 0.051 Fill rate 0.011 0.011 0.010 0.282 0.028 Inventory 0.011 0.011 0.010 0.282 0.028 2 nd and 3 rd iteration: 0.288 0.147 0.500 0.066 Final O / M ratio Healthy Safe Tools & equip. Result Accidents 0.045 0.119 0.310 0.032 0.187 MSI 0.201 0.271 0.140 0.083 0.173 CSI 0.066 0.271 0.140 0.018 0.130 Absenteeism 0.021 0.119 0.205 0.012 0.126 Overtime 0.287 0.036 0.035 0.012 0.106 MTTR 0.153 0.089 0.059 0.158 0.097 OEE 0.080 0.050 0.073 0.112 0.074 Backload 0.124 0.025 0.019 0.032 0.051 Fill rate 0.011 0.011 0.010 0.311 0.030 Costs 0.011 0.011 0.010 0.230 0.025 PEOPLE PILLAR RESULT 1 st, 2 nd and 3 rd iteration: 0.452 0.158 0.034 0.092 0.263 Final Motivation Management Employment planning Communication Work environment Result MSI 0.282 0.284 0.170 0.281 0.205 0.258 OEE 0.215 0.112 0.297 0.154 0.042 0.150 WG status 0.159 0.058 0.017 0.207 0.021 0.106 MTTR 0.082 0.217 0.219 0.083 0.059 0.102 Accidents 0.010 0.012 0.015 0.030 0.293 0.086 CSI 0.024 0.080 0.119 0.062 0.157 0.075 Ideas implementation 0.113 0.020 0.010 0.114 0.012 0.068 Overtime 0.056 0.030 0.056 0.015 0.080 0.055 Absenteeism 0.038 0.028 0.027 0.009 0.116 0.054 MTBF 0.020 0.160 0.072 0.045 0.014 0.045 96 References 1. P. Willmott, ?Total productive maintenance. The Western way?, Boston MA, Butterworth-Heinemann, 1994. 2. S. O. Duffuaa, A Raouf, J.D. Campbell, ?Planning and Control of Maintenance Systems. Modeling and Analysis?, NY, John Wiley & Sons, NY, 1999 3. R. S. Kaplan, D. P. Norton, ?The balanced scorecard: translating strategy into action?, Boston MA, Harvard Business School Press, 1996 4. M. Modarres, S. W. Cheon, ?Function-centered modeling of engineering systems using the goal tree?success tree technique and functional primitives?, Reliability Engineering and System Safety, Vol. 64, Issue 2, May 1999 5. T. L. Saaty, ?The analytic hierarchy process?, NY, Mc Graw-Hill, 1980 6. D. Mather, ?The Maintenance Scorecard: Creating Strategic Advantage?, NY, Industrial Press, 2005 7. M. Modarres, M. Kaminsky, V. Kristov, ?Reliability Engineering and Risk Analysis?, NY, Marcel Dekker Inc., 1999 8. R.D. Palmer, ?Maintenance planning and scheduling handbook?, NY, Mac Graw- Hill Handbooks, 1999 9. T. Wireman, ?Developing performance indicators for managing maintenance?, NY, Industrial Press Inc., 1998 10. B.W. Niebel, ?Engineering maintenance management?, Second edition, NY, Marcel Dekker Inc., 1994 97 11. R.F. Pagano, ?An organization tool to enhance work motivation - part I - Job Satisfaction? Physician Executive , 1993 12. C. Ray Asfahl, ?Industrial Safety and Health Management?, NJ, Prentice Hall, 2003 13. T. L. Saaty, ?The Analytic Network Process: Decision Making with Dependence and Feedback?, , Pittsburgh, PA, RWS Publications, June 2001 14. A. Ishizaka, D. Balkenborg, T. Kaplan, ?AHP does not like compromises: the role of measurement scales?, University of York publications, September 2005 15. C. D. Wickens, J. G. Hollands, ?Engineering Psychology and Human Performance?, NJ, Prentice Hall, 2000 16. M. Bevilacqua, M. Braglia, ?The analytic hierarchy process applied to maintenance strategy selection?, Reliability Engineering & System Safety, Issue 7, March 2000 17. A. W. Labib, R.F. O?Connor, G.B. Williams, ?An effective maintenance system using the analytic hierarchy process?, Integrated Manufacturing Systems, MCB University Press, 1998