Chapter
Backlog Management
ontrol of backlog is fundamental to successful management of the maintenance function and thereby to realization of reliability. It must be held within manageable control limits by preserving a balance between resources and workload. Backlog is defined as the net workload, measured in labor-hours, requested but not yet completed.Another definition is that the backlog is all identified work remaining to be performed. If a job has been started, only the portion of labor-hours still to be completed remains in quantified backlog. Some of the jobs in backlog are delinquent in terms of promise date, but most are not delinquent. Being part of backlog does not connote delinquency. Maintenance work is continuously created over time as:
Equipment is utilized and reliability diminishes Facilities are exposed to weather and usage PM/PdM inspections and other routines come due Operational processes are altered or replaced Planning for Maintenancemeliability Excellence begins with Macro-Planning, which is perpetual balancing of maintenance resources with maintenance workload. If balance is not preserved between these two variables, the operation cannot achieve reliability. Misuse of assets and deferral of essential maintenance work accelerates deterioration, further increasing the requirement for maintenance resources. When workload exceeds resources on an ongoing basis, the maintenance organization falls behind the rate of deterioration and the
69
70
Chapter 7
amount of reactive maintenance grows. You can walk through a plant in this state and see the deferred jobs, broken and rusted items, and general disarray that result when deterioration gets the upper hand. This situation is common and deleterious to morale, product quality, safety, and job fulfillment for maintenance workers. When an organization fails to stay ahead of deterioration, it cannot maintain reliability and certainly cannot climb out of an existing reactive condition without influx of adequate resources. Without such influx, it is impossible for the facility to be proactive because response to failures detracts from resources available for backlog relief, PMPdM inspections, and newly-identified corrective work continually adds to backlog. The resource-demand imbalance continually worsens! When an organization is in this state, any deficiencies found by proactive inspections are not likely to be corrected before failure. This condition renders proactive crews (PMPdM and Backlog Relief) impotent, destroying morale and effectiveness of the entire maintenance organization. Think about that! Inspections are a proactive investment. Prompt correction of identified deficiencies provide the return (ROI). Failure to capture the return on investment is asinine. The backlog assigned to each crew or trade (measured in weeks) must be calculated and displayed on a monthly basis. The calculation vehicle is the Maintenance Work Program discussed later in this chapter and the next. Smart managers use this vehicle to base their decisions upon facts. Decision areas include: Staffing by type of work and skill required. Maintaining required resources by adjusting overtime, contrac tor support, or staffing; in that order of flexibility. Evaluating capacity available to handle project work and/or major shutdowns. Deciding if and when to utilize contract support to meet abnormally high demands.
BACKLOG MANAGEMENT
Well-managed maintenance departments hold backlog within established upper and lower control limits. Backlog Ready to be Scheduled is part of but isolated within Total Backlog. Two-to-four weeks
Backlog Management
71
of Ready Backlog and four-to-eight weeks of Total Backlog are commonly accepted control limits. Ready Backlog is just that-jobs that are ready to be scheduled. All tools, parts, materials, drawings, and authorizations are in hand. The job can be scheduled any time Operations can release the asset to Maintenance and required maintenance resources can be made available. During the Weekly Coordination Meeting, the parties agree upon a specific day and time to schedule each job. Total Backlog includes Ready Backlog plus all the other open work orders still in one or more Status Holds. These latter jobs could be missing parts, authorization, budget, or some other element (see Figure 7.1). Too much backlog means excessive delay in response to customer needs and ultimately leads to Deferred Maintenance, the dreaded D word. Deterioration gains the upper hand. Capacity, reliability, safety, and customer service suffer. People requesting lower priority jobs notice that their jobs never get scheduled. The resultant lack of customer satisfaction contributes to the low esteem in which the maintenance function is held in many organizations. A well-designed and administered Priority coding structure with an aging feature can prevent such jobs from becoming indefinitely delayed or lost in the backlog. Even medium- and low-priority jobs must be completed within a reasonable time frame.
Job Control
Backlog is made up of sub-groups
n\
Schedule
TOTAL BACKLOG
Waiting for Materials
1
,
WISH LIST
Figure 7.1
72
Chapter 7
Backlog is the holding tank that helps regulate the flow of work into the maintenance shop (Figure 1.2 and 7.2). Jobs are scheduled in a smooth flow limited by available labor-hours. When Ready backlog increases beyond four weeks or Total backlog beyond eight weeks, the consequences can be severe (the holding tank overflows).
( I
Backlocl is the key to control
Evens out flow into shops Determines status and what is missing Ready verses Total
1'I
I1
Metric: Trend Ready backlog+md Total Backlog measured in weeks
Figure 7.2
Too little backlog indicates insufficient workload to keep the existing maintenance staff effectively deployed and utilized. Work tends to be stretched to fill out the day and crucial maintenance resources are wasted. Backlog quantification (derived from Work Programs) should be regularly compared to available labor resources. Resources should be flexed (adjusted) up or down as analysis indicates necessity. There are three flexible resources for consideration.
1. Overtime is the easiest resource to adjust, assuming that it is not already excessive. An overtime range between 7% and 15% of paid straight time is recommended. Too little overtime is not sound maintenance strategy. Too much overtime is not healthy, safe, or productive for either the company or the employees.
2. Contract Support is the second most flexible resource.
3. Permanent Staff is the least flexible alternative. This form of adjustment should be deployed only when resources and workload
Backlog Management
73
are outside the target range for backlog weeks, no relief is in sight, and the condition appears to be permanent (or at least semipermanent).
EXAMPLE OF BACKLOG OUTOF CONTROL
At a steel mill in Western Pennsylvania, the total backlog was 96 weeks per tradesperson! That meant that, even if the plant were shut, the current resources would need almost two years to clean up the backlog. The consequence was a complete lack of management control. Additionally, there was a nasty work environment with low morale and self-esteem, a poor safety record, poor customer service, low profits, and a constant state of fatigue.
1
JOB STATUS
It is essential that plannerhchedulers know the current status of each job in the backlog. Such status is documented via Status codes within the Computerized Maintenance Management Information System (CMMIS). The Status Code feature brings order out of chaos. It is the most important code in the entire Work Management System (WMSEMMIS), especially for Planners/Schedulers. An example of status code structure is presented in Figure 7.3 on the following page. The existence of a coherent backlog list by status answers questions generated fkom all levels of the organization. The CMMIS must enable planners to access backlog by:
0 Planner supporting the area, trade, or job
0 Job status
0 Crewhpervisor
0 Asset, machine, or production line
0 Originator
0 Age of work order
0 Due date
74
Chapter 7
Form of backlo
Major shutdowns several months off are held in an independant backlog
Figure 7.3 Work Order Status Codes
Each of the access paths to backlog answers different questions from different questioners. For example, sorting the backlog by machine identifies other jobs that might be performed during the access window to the given asset. Backlog must be complete, current, pure, and reliable. Trends of Backlog Weeks are going to be presented to senior managers with a recommendation for force adjustment of one form or another. The basis of recommendations better hold water. How would the integrity of your current backlog stand up to the test shown in Figure 7.4?
Backlog Management
Does your backlog include.. .?
a a
75
Work Orders that are completed, but nobody has bothered to close them out? Duplicatejobs under different names or different work orders Jobs greater than a year or two old-excluding those awaiting a known, hture shutdown? Jobs for which no one recognizes the originator or why the job was needed in the first place Smalljobs supportive of larger jobs, but not recognized in backlog Poorly describedjobs (no one knows what to do, or what to do it to) Job status not recorded or not updated and nobody recalls current status. Were parts required? Were they ordered? Were they delivered? If so, where are they now? Jobs needing to be done that are not recorded in backlog
Figure 7.4 Checklist for Validation of Backlog Integrity
A few suggestions regarding the backlog validation process are offered: Sort the backlog by asset or work center. Involve the asset custodian to participate in the validation process. When removing duplicate jobs, remove the lower priority or the most recently requested. If the estimated labor-hours differ, re-estimate them. Transfer all appropriate detail to the work order remaining in backlog.
Do not eliminate aged jobs without determining if they are still required.
Needed jobs beyond current budget must be brought to management attention with a case for approval. If there are two or more work orders for the same PMPdM routine on the same asset, the oldest one should be closed with notation that it was missed (non-compliant). This effort is not necessary in a well-designed CMMIS because a shadowing feature catches the duality and automatically closes the oldest. It is important that the non-compliance be captured in
76
Chapter 7
Equipment History so that P W d M Effectiveness in relation to Reliability can be accurately evaluated." Jobs not captured in the backlog constitute a special problem. Internal customers may be conscious that Maintenance is already overloaded and see no point of compounding the situation. As a result, they simple stop inputting to the Work Order System, except for failures (urgent work). The reality is if the work needs to be performed, Maintenance must be made aware of it. Such work must be evaluated relative to other demands and assigned a prioritykriticality ranking. An ignored maintenance problem might catapult to the forefront due to an accident, a production increase, a quality initiative, or ergonomics. If the job is not added into a controlled backlog, Maintenance does not know of it-at least officially. If a need exists, it should be identified and quantified as part of the backlog. Otherwise, it simply swirls around in the background as part of the amorphous mass of undefined deferred maintenance.
