Effective Backlog Management – Backlog Size Control
Backlog management has a number of different but, interdependent focuses: Backlog Work Order Quality, Age of Backlog and Backlog Size Management. This article will focus on Backlog Size Management. In parts 2 and 3, the Age of the Backlog and Backlog Size Management were discussed in detail.
During the 1960s many major companies reduced crew size by laying off junior workers, rolling maintenance employees to operations roles or temoprarilly laying them off. It was understood during that period, that a newly hired employee could expect several temporary layoffs until they gained enough seniority to be above the layoff threshold. Major companies realized the shortsightedness of this practice in the mid seventies, and began using other means of adjusting the crew size when business was slow.
A common approach during the mid-1970s was to staff at a 60-80% level and maintain a supplement contract work force to meet the requirements during high demand periods.
I experienced this new approach in the early eighties when the plant where I was assigned encountered an extended slow business period. The company reacted first by displacing all contractors with their employees – security, janitorial, and supplemental maintenance. As the slow period became extended, mechanics and operators were loaned to local community service programmes, such as Habitat for Humanity. For several months, I had mechanics building affordable housing while being fully paid by the company. Needless to say, the morale and loyalty of the employees grew tremendously. The company also gained from reducing turnover and training costs.
Reactive organizations will have very large backlogs (documented and/or undocumented). This may be interpreted as an indicator of understaffing. In reality, adding resources will not have a substantial impact on the backlog size without fundamental changes in how maintenance tasks are addressed and the quality of the overall maintenance programme.
The nature of a reactive maintenance programme will keep the focus on the emergencies of the day, causing less urgent work to be ignored until it becomes one of the next day’s emergencies. With this firefighting mentality, short cuts and Band-Aid maintenance techniques become the norm, guaranteeing more emergencies and shorter life cycles. This cycle will continue until a strong effort to implement a proactive work management process is taken.
High level metric of staffing levels
Organizations with a proactive programme and a well-implemented work management process can use backlog size as a high level metric of its staffing levels (overall and individual craft mix) and the quality of its Planning and Scheduling process.
This is not a metric that triggers rapid action, but should cause a thorough investigation to determine the root cause. World Class maintenance programmes have a 5-week backlog target by craft and crew with a 3 to 7 week upper and lower control limit. This should be monitored on a monthly basis, but only trigger an investigation if a trend is established over a number of months. Once a trend is recognized, an objective investigation should be undertaken to determine why the trend is occurring. There are normal activities that could cause a temporary trend in the backlog level. For example a major plant outage or turnaround can push the backlog higher when many routine tasks have to be delayed.
Reaction to the trends can result in cleaning the backlog (see Part 1 of this article in Maintworld 3/2017) or reducing/increasing the contract crew size to bring the backlog within the control limits. Today, many companies use supplemental maintenance as a relief valve during difficult business periods.
Some companies take a long term view, allowing reduced replacement of attrition to lower the crew size to match the workload. In cases of a high backlog, both increasing overtime and bringing on more contractors are commonly used tools.
A note of caution if you allow your supplemental contractors to maintain their own backlog: any decisions to reduce or increase crew size must be done only after a thorough review of the quality of the work orders in backlog.
I experienced the need to take this step a number of years ago. The supplemental maintenance contractor manager requested increasing the number of his staff due to a high backlog. His request was at first considered and he started the process of hiring more supplemental mechanics. When knowledge of this increase reached the maintenance planners, a number of questions were raised. A thorough review of each work order revealed many with unreliable estimates, work orders that had already been completed or work orders for tasks that had no business benefits. The contractor’s backlog was reduced to normal levels and the hiring of additional supplemental mechanics avoided. To ensure this didn’t reoccur, a regular backlog review meeting was established.
Contract firms do use backlog metrics as a means to control job security and profitability for the firm by moving personnel between jobs, if possible, or reducing crew size (layoffs) to match the available work. They will set the backlog targets at an appropriate level for the industry they are serving, normally much higher than the 5-week target of world class companies. Specialty contractors can have much higher backlog targets, especially if there is a limited number of firms specializing in their field.
Establishing a Backlog Metric
A clear definition of when a work order is placed in backlog should be developed first to ensure an accurate backlog. Some organizations include all work orders, even those without accurate estimates. This adding of work orders with “guesstimates” will make the backlog greatly exaggerated.
To ensure an accurate backlog, only those work orders that have been fully estimated should be included. Work orders that are awaiting planning or being planned should not be added to the backlog.
Any backlog size metric will only be as accurate as the planner's estimates. World class metrics for work order estimate accuracy is =/- 10%. Reactive maintenance organization’s work order estimates will be +/- 50% to many times more.
A separate backlog metric should be kept for each craft and crew (multi shop sites) with control limits to trigger any possible actions.
While the man-hours per craft and crew are what are pulled from the work orders, converting the metric to man-weeks helps make the metric easier to use.
Typical Backlog metric
This metric should be reviewed on a monthly basis and investigated when a trend of several months is seen. The first step should be a quality review by those personnel who are familiar with the operations needs and maintenances resources (Operations and Maintenance supervision or managers). This review will need to look at all the work orders in backlog, not just those that are overdue. If the quality review doesn’t bring the backlog into control limits, a continuous improvement tool such as a 5 Why analysis can be applied to discover the cause of the out of control backlog.
Possible causes of an out of control backlog are:
- Work order quality – inaccurate estimates – invalid work orders – completed work
- Major turnaround being done with plant maintenance or supplemental maintenance personnel
- A time-driven capital project being completed with plant maintenance or supplemental maintenance personnel
- Business slow period – short or long term
- Crews not correctly sized to the workload
- Crafts not sized correctly – too few or too many of one craft
- Inexperienced mechanics
Corrective steps that could be taken:
- Clean the backlog
- Be patient and let time correct the cause (turnaround, project, or a temporary business slowdown)
- Balance crews across all shops
- Identify areas of inexperience and provide training
- Reduce or increase the supplemental maintenance personnel
Regardless of what the cause is determined to be, any corrective steps should be carefully considered for their long-term effect. Any corrective effort should be discussed with all stakeholders and a consensus reached.
A growing number of companies want to use big data analytics in their predictive maintenance and are also investing in the resources needed for this. Of the companies already using this technology, no less than 95 percent say that they have already achieved concrete results. This is the conclusion of a follow-up study conducted by PwC and Mainnovation among 268 companies in the Netherlands, Germany and Belgium.
In today’s operating and production environments, systems and equipment must routinely perform at levels that were not possible a decade ago and which were unthinkable thirty years ago. Requirements for increased availability, throughput, product quality, agility, and operating effectiveness within a rapidly-changing demand environment continue to elevate the tempo and intensity of operations.