Maintenance is the business of managing asset defects. Day to day, this business starts with defect detection, documented in a notification. This critical initiating step kicks off the process of work management to approve, plan, schedule, and execute the work. Notifications are the fuel of this process, but due to a lack of clarity, the notifications often leave the engine sputtering. By standardizing and improving the information content in notifications, you can set the stage for better performance at each stage of your work management process.
Most work management improvement projects focus on planning and scheduling, giving notifications little attention. Instead of rejecting insufficient notifications and coaching the notification writers, supervisors and gatekeepers run ragged researching, troubleshooting, and rewriting to make the process work. This not only enables continued poor notification quality, but it also shields management from the reality of the problem: bad notifications are crippling the entire work management process.
It does not have to be this way. You can make a lasting change in how your organization writes and uses notifications with a three-pronged approach:
- First, notifications should be standardized and employ simple, structured menus of operational failure modes tailored to the selected equipment type. This provides structured data for routing, prioritization, and analysis. It also provides those who initiate notifications a frame of reference that will enhance the quality of their narratives.
- Second, a cultural change in how people view the ownership and importance of notifications should be the foundation of the training on and implementation of this tool. Notifications must be seen as a reflection of individuals’ sense of ownership and their professional knowledge of the systems with which they work.
- Finally, these changes must be monitored and supported through quality reviews, coaching, and management attention that carry them through until they become sustainable practice.
The time and attention invested in this notification improvement effort will pay dividends that can be monitored and quantified:
- It will lower “rework” time on the part of high-value gatekeepers and planners.
- It will shorten processing time, resulting in quicker execution of identified work.
- It will result in better plans and more effective repairs.
- It will enable the organization to move toward standardized maintenance plans, thus leveraging the repeatable nature of maintenance work.
Poor notifications cripple the maintenance process
Maintenance work management is a serial process: notify, review/approve/request, plan, schedule, coordinate, execute and, finally, verify work quality. For the vast majority of maintenance actions, a notification should travel through this process one way, one time, quickly and efficiently. This is not the reality in many plants.
To start with, notifications are literally a blank screen with little or nothing to guide the person responsible for submitting the notification. Too often, this results in no more than a few words describing the problem: “Pump broken.”
Vague notifications don’t provide enough information to sail through the work management process. It will circle back again and again, wasting valuable time for all involved. In some cases, a poor notification can result in a lost opportunity to execute a maintenance intervention before functional performance of the asset is adversely affected.
In the absence of a well-written notification, gatekeepers are forced to hunt down shift supervisors and board operators in an attempt to decipher the vague description. These stakeholders end up scratching their heads trying to determine what was meant by the individual who submitted the notification. Approvals will be delayed. Priorities and other details are reduced to guesswork. The planner will walk circles trying to understand where the leak might have been before they switched to the B pump and blocked in and bled the A side. The plan will be a best guess. The scaffolders, too, will be in the control room asking about this notification, since their job step reads only, “See operations for location.” Finally, the millwrights will have their crack at it. Hopefully, the leak was truly in the pump and not the flange that they were told did not need replacement.
The all-too-common poor notification is not a harmless trifle. It piles extra work on people at each step of the process. At T.A. Cook, we have repeatedly documented instances where highly compensated gatekeepers and planners have spent a quarter of their day running down poor notifications. Furthermore, an insufficient notification is the gift that keeps on giving back to operations in the form of repeated questions, field walks, and – ultimately – delayed and ineffective repairs. These all have very real costs in the form of reduced employee productivity and increased equipment downtime. There is a better way.
Simple standardization goes a long way
A notification is meant to identify and communicate a defect – an actual or impending operational failure condition – requiring maintenance action. This central component of a notification is left to chance, however. The blank canvas of a free text field does not provide the structure needed for success. No combination of training, instructions, reviews, and cajoling is going to overcome the power of this blank space to ruin data quality. Consistently good notifications require standardized failure mode choices in the form of a menu.
This use of “structured data” gives users a starting point and a framework of “buckets” in which to group and categorize their notifications. Structured data enables data analysis over time, as well as a consistent and concrete way to categorize notifications for gatekeeping and planning. It also helps users create a better narrative description in the free text fields by giving them a starting place to guide their diagnosis. For example, health care professionals have found that the Electronic Health Records doctors use to document and communicate their diagnoses are most effective with a similar combination of structure and free text fields.
The good news is that most computerized maintenance management systems (CMMS) already have a failure mode field that can be customized to fit your needs. Thus, any attempt to standardize notifications must start with a manageable list of failure modes. The menu should cover the roughly 80% of issues that we see repeatedly in the normal universe of routine maintenance.
ISO 14224 and IEC 60812 provide comprehensive lists of failure modes as a starting point. In practice, we recommend simplifying these lists to roughly eight broader categories. Don’t over-engineer your taxonomy of failure modes for the notification process. We’re not trying to automate the root cause analysis process.
Furthermore, the menus should be customized by the selected equipment type. A single menu would have too many options – and many that do not match the selected equipment. On the flip side, we cannot have menus customized to every pump type, model, and manufacturer. A good approach would be to create customized menus for the 23 equipment classes found in ISO 14224, noting that not all 23 classes will apply to every plant. Figure 1 provides an example of the resulting menu for a pump.
This approach offers several benefits. First, it provides the notification initiator with a concrete framework in which to work. At a minimum, selecting a failure mode provides more information than “pump broken.” Second, it provides engineers with structured data that can be analyzed to improve asset reliability.
Third, and most usefully, using standardized failure modes to write notifications provides concrete data in order to prioritize work and to identify the proper, repetitive maintenance action, which can be standardized to improve work management efficiency and effectiveness. Notifications can be explicitly linked to the appropriate standard job plan.
Write it like you own it
Standardization alone will not change how employees view and use notifications. It is important that they see notifications as a reflection of their sense of ownership and systems knowledge. Too often, this is not the case. More experienced technicians and operators often do not take the time to write, review, or coach on notifications. The task is left to newer hands who are not imbued with a sense of the value of a good notification and often lack the experience to write one.
We would act differently if it was our prized, classic (pre-onboard diagnostics) pickup truck and not pump 123A. We would be sure to troubleshoot and mitigate the issue. If we could not go in person, we definitely would not send our twentysomething son or daughter down to the garage without detailed instructions on how to describe the issue. We might even write it down – proud of our systems knowledge and protective of our asset. In this, we would take time to teach our son or daughter about the truck and show them the right way of doing things. When we talk about having a sense of ownership and pride, this is what it should look like.
Maintenance and operations leaders must view an equipment malfunction as an opportunity to correct a defect properly – or even better, to understand and eliminate a recurring defect. It is a real-world opportunity to review and reinforce systems knowledge. Moreover, it is an opportunity to display and enhance the whole team’s sense of ownership and professional knowledge by conducting well-informed diagnosis and troubleshooting and writing it up in a quality notification.
Effective troubleshooting and the proper writing and submission of work notifications must be part of initial training, ongoing coaching and mentoring, and periodic reviews for plant personnel. The importance of coaching, mentorship, and on-the-job training can’t be overstated. It is critical to providing maintenance and operations personnel the tools and the attitude necessary to be good stewards of plant equipment.
This appeal to ownership is a critical element of any attempt to improve notification quality. Efforts to make this change, however, must go beyond motivating phrases in the classroom and in company communications.
Upholding the standards
The initial boost in attention to notifications cannot be sustained by standardization and slogans alone. Improvement takes time – and someone watching the score. Several lines of defense exist when it comes to upholding the standards.
An approval process must be in place – whether in the form of a workflow through shift/shop supervisors, or at the gatekeeper level – that is more than a rubber stamp. Beyond the notification process, this is an important part of on-the-job training. There must be a willingness to send inadequate notifications back to the originator for revision. When a notification is rejected as inadequate, supervision should be involved to reinforce the importance of good notifications and to ensure that the originator has the knowledge and support needed for adequate write-ups.
Additionally, management should monitor this effort with a pair of metrics: rejection rate and notification quality (see Figure 2). Rejection rate can be taken directly from the CMMS workflow. Notification quality should be checked through a monthly audit process. Select a random sample of notifications and score them against specific criteria in categories such as equipment identification, location information, failure mode, description details, and priority.
Set a baseline by checking your rejection rate and audited notification quality at the beginning of the improvement effort. Compare the baseline rejection rate against the audited notification quality and conduct a why-why analysis to determine what is driving rejections and quality issues. You will also want to understand false positive and false negative rates: good notifications that were rejected, and insufficient notifications that were approved. With this baseline, you will be much better able to understand what is behind the numbers and how they can guide your improvement effort. Collectively, this information is much more valuable than rejection rate alone. You will know if the rejection rate matches quality and will have a deeper, quantified understanding of your initial and ongoing opportunities for improvement.
An effective and efficient way to conduct your ongoing audits is to select a random sample of notifications to be reviewed and divide them among a team of experienced operations and maintenance leaders. This makes the workload manageable, as well as improving supervisors’ first-hand knowledge of remaining issues. The results of this monthly audit give a second check on quality – as well as producing a team that will support ongoing improvement.
Resist the temptation to use these metrics as a cudgel, and consider what systemic factors are hampering improvement. Is coaching and on-the-job training actually happening to reinforce your new expectations? Are competing priorities for operators’ time making them rush through the notification? Are control room operators filling out notifications based on a short radio call because field operators do not have easy computer access? Are they facing technical issues? Make sure that such roadblocks are uncovered and removed – and do not assume that because no one said anything, there are no problems. Ask the question.
Finally, offer rewards for well-written notifications that add value by proactively and clearly identifying a defect. Rewards can range from informal praise, to positive performance documentation, to tangible awards similar to “safety bucks” programs. In all cases, the reward must be tied to notifications with safety, quality, or production value. You do not want people stuffing a lottery box with worthless notifications.
Track and communicate the benefits
As your improvements gain hold, it is important to communicate the benefits that the team is producing. It is great to see rejections dropping and quality improving, but two other lagging metrics can help to quantify the end results of this relatively simple effort.
Improved notifications should shorten the processing time from identification through planning, at a minimum, and in some cases can result in better plans that avoid delays and rework. In order to quantify and communicate this success, track processing time from notification to work order creation and work order creation to technical completion. Keep an eye out for good notifications that result in timely corrections of defects in safety or production critical equipment. These can be quantified in terms of incidents or lost production opportunity avoided, as well as providing a winning narrative to maintain support for your initiative.
Conclusion
As these lagging metrics will show, poor notifications hobble maintenance efforts and result in real monetary impacts. Gatekeeping, planning, scheduling, and execution all build on the information found in the notification, so take some time to fix that foundational element. Done right, it can make a quick and significant return on a relatively small investment of time and attention:
- Educate stakeholders on the value of good information content.
- Provide simple, standard failure mode choices, tailored by equipment type, which are to be expanded upon in the description field, or long text.
- Train and coach operators and artisans on the new standards.
- Ensure that there is an appropriate review and approval process in place.
- Monitor the improvement effort via notification quality and rejection rate metrics.
- Track and communicate the benefits to the organization.