Implementing a Computerized Maintenance Management System

Implementing a Computerized Maintenance Management System

11 Implementing a Computerized Maintenance Management System Profound knowledge comes from the outside, and by invitation. A system cannot know itse...

78KB Sizes 0 Downloads 31 Views


Implementing a Computerized Maintenance Management System

Profound knowledge comes from the outside, and by invitation. A system cannot know itself. W. Edwards Deming Most executives wouldn’t dream of running their companies without an effective computerized accounting and financial management system, and yet many routinely run their production equipment without an effective computerized asset management system for the maintenance and reliability (and cost) of their equipment, often valued in the hundreds of millions.1 This chapter describes how one of Beta’s divisions accomplished that task. Beta, like most medium to large manufacturing organizations, has put in place a computerized maintenance management system (CMMS) at most of its plants, with informal surveys indicating that over 90% of its plants have some form of a CMMS in place. However, a closer review of those plants and questions about the specific use of their systems revealed that most— 1. Are not using the CMMS for all appropriate equipment in their plant. For most plants some 25% to 75% of equipment is not in the database and therefore not maintained using the CMMS.




2. Are using the CMMS primarily for issuing work orders and scheduling work, but often not using it for comprehensive planning, for development of equipment repair and cost histories, for performing Pareto analyses to help prioritize resources, etc. 3. May be poorly trained in the systematic use by all the appropriate staff of all the capabilities of a good CMMS. 4. Don’t fully apply many of the other tools available, e.g., stores and purchasing interface, resource allocation, document retrieval, etc. Many times, someone was told to go buy a system and use it—a kind of magic elixir for the ills of the maintenance department. This type of behavior reflects a symptomatic response to maintenance issues vs. a systematic approach, fairly typical with most manufacturers. The logic seems to be “If we just do . . ., then everything will get better.” You fill in the blank with the flavor of the month. It’s much like treating a cough with cough drops so the cough goes away, rather than understanding that the patient has pneumonia and requires long-term rehabilitation. Few companies at the beginning of a CMMS implementation effort recognize that a CMMS is no more than a very sophisticated software shell, which is empty at the beginning—users must fill the shell with data and then train most of their staff to ensure its effective use. Beta was no exception. Another way of thinking about this would be buying a word processing program and expecting that books would be more readily developed. Certainly they could be, but the word processing software is just a tool to facilitate the creative process. In fact, the way many of Beta’s maintenance departments were using their CMMS, they could achieve the same result with a word processing package, which usually has a calendar to schedule and issue work orders. This is not recommended, but it serves to highlight the effectiveness of many CMMSs in use. Something is wrong with this, considering the millions that are being spent for CMMS programs. The fault lies not in the vendors, so much as in the expectations and implementation process of the users. Users must understand that a CMMS is only a shell, which must be filled with data and then used in a comprehensive way for effective mainte-


nance management. A CMMS won’t solve equipment reliability problems any more than an accounting system will solve cash-flow problems. Both are systems that must be filled with data, data that are subsequently analyzed and used to good effect. A good CMMS is most effective when used for: 1. Work order management—scope, effort, trades, procedures, permits, schedule, etc. (more than just issuing work orders). 2. Scheduling of essentially all work, and planning major efforts, parts, tools, procedures, skills (not just scheduling). 3. Equipment management, especially cost and repair histories; and bills of material for each piece of equipment and their related spare parts. 4. Use of equipment histories for Pareto analyses, prioritization and allocation of resources, and identification of key equipment needing root cause failure analysis. 5. Purchasing and stores management for spares, “kitting,” planning, and reliability improvement. 6. Document control and ready access to manuals, schematics, and procedures. 7. Resource allocation for effective reliability, production, and cost management. Moreover, preventive maintenance in the classical sense of fixed interval tasks is typically nonoptimal—How many of your machines are truly “average”? How often are you over- or undermaintaining using a mean time between failure approach? As we’ve seen, the data, when available, typically suggest that less than 10% of most equipment life is near the average life for that class of equipment. As noted previously, preventive maintenance is best: 1. When used in conjunction with a strong statistical basis, e.g., low standard deviations from average for a strong wearrelated failure mode; and when condition assessment is used to validate the need for the PM. 2. When condition assessment is not practical or cost effective.




3. For routine inspections and minor PMs, but even then could be optimized. 4. For instrument calibrations, but even then could be optimized by tracking calibration histories. 5. For some manufacturer PMs—that is, those the manufacturer can back up with definitive statistical data. 6. For some staff-defined PMs—likewise, those backed up with definitive statistical data. 7. For regulatory-driven requirements, e.g., OSHA 1910, 119j, code requirements, etc., but even then could be optimized with histories and effective condition monitoring. One last point before proceeding with Beta’s case history—a CMMS is no more than a tool, and tools don’t provide discipline— processes do! A disciplined maintenance process must be put in place for the tool to be effective, not vice versa.

Case History With this in mind, this section provides a case history of how Beta International’s specialty products division (SPD), with limited previous experience in computerized maintenance management systems, effectively implemented its CMMS to help ensure lower maintenance costs, higher staff productivity, and improved equipment reliability. SPD was faced with an extraordinary set of problems—reduced revenues, increased costs, lower capital and operating budgets, and a maintenance department with: 

Expensive breakdowns

Little planning and scheduling

Excessive forms, paperwork, and bureaucracy

Limited equipment histories

Superficial replacement/repair criteria

Minimal equipment failure analysis


Ineffective resource allocation criteria

Limited control of maintenance costs

Limited measures of performance or effectiveness

Or, as the saying goes, “Other than that, they were pretty good.” Clearly, something had to be done to improve this situation, and it was viewed by the company as a major “opportunity” for improvement. A part of the solution to this opportunity was to develop a more effective maintenance function, one that provided greater control, improved reliability of the equipment, and promoted greater teamwork between maintenance and operations. The first decision made was to put in place a computerized maintenance management system—a fairly typical step. However, to ensure the effectiveness of this decision, SPD put together a strategic plan that defined: 1. A process for CMMS selection 2. Potential impediments and their resolution 3. A process for implementation 4. A method for measurement of results 5. A continuous improvement process that would ensure successful implementation beyond initial successes and that would not just be viewed as another “magic bullet”

Preliminary Authorization and Selection Initially, management gave tentative authorization to select and purchase a CMMS. Prior to making this decision, the selection committee made several site visits to review various vendor products in a real working environment. Its conclusion from this review was that it didn’t know how to effectively select a system. This led to hiring a consulting specialist for the selection team to facilitate the specification of the CMMS that would meet its needs. After considerable effort defining its needs and processes, the team finalized a specification and put forth a request for proposal (RFP) to several vendors.




After reviewing the proposals, the team narrowed its selection process down to a few vendors that performed office demos of their products. From there the list was narrowed even further by performing site visits, without the vendor, to validate vendor representations, to see the systems in action, to use the systems in a working environment, and to ensure that the system they might choose was effectively employed by others and would meet the team’s requirements.

Selection Criteria The selection process and criteria went beyond the site visits and vendor demos. Specifically, the selection criteria included: Value: 1. Conformance to the RFP (including a client/server system architecture) 2. Features and functions (including using the system at a working site) 3. User-friendly (or hostile) characteristics (using the system at a working site) 4. Vendor support systems (including calling as a customer and evaluating the support) 5. Vendor training (including attending a training class before purchase) 6. User satisfaction with the system/results (vendor’s five, plus a random selection) 7. Vendor customer base (total number of customers served) 8. Vendor financial strength Cost: 1. Initial cost 2. Continuing costs—training, upgrades, phone support, etc.


Primary decision criteria: Value took priority over cost, so long as cost was within a reasonable range. Low bid was not the criterion. Finally, it was found that the actual implementation cost was approximately four times the initial software cost. Managers take note: The implementation process is considerably more expensive than the initial cost and should be considered in the decision-making process.

Impediments and Solutions There were several impediments to the project from the beginning, each of which was addressed effectively: 1. Culture of the organization: a. “If it ain’t broke . . .”; or its brother—“We’ve always done it this way.” b. Prospective job change requirements with the union. c. “You can’t plan maintenance.” 2. Management support and resources—staff and capital Clearly, from the discussion, “it” was broken, and with the loss of revenue and climbing costs, something had to be done. Other efforts were ongoing to improve market share, develop new technology, improve operations, etc. And, this effort was part of an overall effort for improved maintenance practices, which necessarily included predictive and proactive practices. The union was approached with the concept of using a CMMS for solving inherent system problems, to make jobs easier, to reduce the time spent at the parts counter, to have equipment isolated when it arrived, to have the tools available, etc. Reduced revenues, climbing costs, and “broken” processes were all basic drivers for change in the organization. Reluctantly, but with assurances that the process was not targeting head count reduction, management proceeded with implementing a CMMS. Any head count reduction requirements would be managed using attrition, reduced contract labor, lower overtime, and reduced contracting of small capital projects. Performance measures were finally concluded to be the only effec-




tive way to measure the success of the implementation process, but only after considerable discussion and negotiation. One fear the union had was that the system, once implemented, would be used to “bid out” all the work to contractors, because it was now all defined in the CMMS. Union leaders were assured that this was not the case, and they required routine reassuring of this. This effort required a continued fostering of trust between union leadership and management. As noted, “it” was broken. Management was convinced that something had to be done and authorized the initial study to determine what, how, when, who, where, how much, potential benefit, etc. Following this study, it authorized proceeding with the principal project to specify, purchase, and install the system in a pilot plant, which would in turn be used as the model for the remaining six plants, a total of seven plants having the CMMS installed. More importantly, however, it was recognized that simply installing a system would not be effective without a systematic plan for implementation and continuous improvement, and achieving acceptance by the union and shop floor of the CMMS implementation process. Union support would be needed, because it was responsible for much of the data entry, work orders, repair codes, closing of jobs, etc. To that end, the following was set up: 

Implementation at a pilot plant

User group, consisting of a representative from each plant

Newsletter—announcing purpose, successes, failures, actions

“Continuous” periodic training to update learning

Routine meetings to address gossip, rumors, perceptions, problems

Staged approach for easing the effects of change

Implementation committee for facilitating implementation

Pilot Plant Implementation One of the first steps was to select a pilot plant for implementation. After much consideration, and developing the confidence that the implementation would be successful, a mid-sized plant that had


processes similar to its largest plant was selected. This plant was more manageable and would permit proving the process and technology prior to full implementation. At the same time, there was one representative from each of the other plants joining the implementation committee from the beginning to ensure that other plant needs were met, and that when the representatives began the implementation process, they would have a first-hand understanding of all the major issues that had arisen at the pilot plant. The representative, in fact, would be responsible for implementation at his or her plant. Initial expectations and goals as to the timing of the implementation, the training required, the potential cost reduction available, and initial schedules for achieving these goals were established. Startup activities included: 1. A complete inventory of equipment 2. Definition of initial PMs and procedures 3. Solicitation of departmental support from management and skilled trades 4. Routine informational meetings Routine informational meetings were very important to explain the objectives and implementation process; to address any rumors, gossip, or misperceptions; to explain the training that would be done; and so on. Rumors and gossip could take on a life of their own and needed to be addressed quickly and definitively. The consulting specialist continued to facilitate the implementation process and to serve as an outside sounding board and objective mediator on issues of concern. Committees were created to discuss major issues, to back up informational meetings with written word about plans and processes, to discuss successes and “failures”—including the course of action related to any failure—and to generally inform all employees of the system.

Procedures PM procedures were developed using:




Equipment manuals

Interviews with appropriate staff

Discussions with vendors

Better definition of work requirements

Better application of predictive technologies

Better definition of staffing and material/parts needs

Purchasing software was linked to the CMMS and included stores management. Material planning and an automated pick list and kitting were made a part of the work order process. Equipment Database. The criteria used to determine whether or not a particular piece of equipment was to be put into the CMMS database were: 1. Is a history of the equipment needed? 2. Is a PM required for the equipment? 3. Will an operator be able to readily identify the equipment? Based on these criteria, some 20,000+ pieces of equipment were put into the database for all plants. At the largest (and oldest) plant, some 8,000+ items were put into the database over a 2-year period using approximately 8,000 to 10,000 labor hours. Use of the System. Skilled trades routinely did their own time reporting, entered work requests directly, entered closing remarks, and closed jobs out. Backlog was reviewed routinely and used to manage trade work assignments, grouping jobs, planning resources, etc. Management routinely used the system to review equipment histories and to generally manage maintenance activities. Performance measures included backlog, completed work orders, overtime, aged backlog, work order status, work schedule status, actual/planned hours, PM effectiveness, maintenance cost by area, average cost per work order, contractor costs, etc. Other Issues. In retrospect, the selection of a client/server system architecture was essential to the successful implementation of the system, allowing for sharing of data systemwide. However, the ability to integrate various software packages effectively was more diffi-


cult than originally anticipated. This will be reviewed in some detail as part of any future systems implementation process.

Results The results were remarkable. When combined with other efforts, such as predictive and proactive methods and improved operational practices, they achieved: 1. Acceptance and use of the system by the skilled trades, by management, and by engineering 2. Better prioritization of work, scheduling, planning, backlog management, and resource allocation 3. Improved equipment histories for better repair/replace decisions, root cause failure analysis, accounting of costs by process and by machine type 4. Improved accountability at all levels where they were measuring performance 5. Improved stores management and lower inventory levels, and improved equipment reliability 6. Much more effective maintenance—better procedures, reduced paperwork, reduced reactive maintenance through preventive maintenance, and elimination of standing work orders (which had been a black hole for money) The pilot plant became the plant with the lowest unit cost of production among the seven plants: 1. Labor hours per job were reduced by 25%. 2. Job completion times were reduced by 20%. 3. Contract labor was reduced by 60%. 4. Work order volume was actually up 35%, with the elimination of standing work orders.




Total Cost Savings Amounted to Well over $6,000,000 Clearly, effective implementation and use of a computerized maintenance management system can provide extraordinary benefits to a company. To achieve these benefits, however, a systematic process must be implemented to ensure proper system selection, implementation, and use. This process also must be combined with predictive and proactive methods, and must fully integrate the engineering, operations, and purchasing functions.

Reference 1. Moore, R. “Implementing a Computerized Maintenance Management System,” Reliability, Directory Issue, 1996.