Savvy Data Decisions

Editor’s note: This article is the first part of a two-part series on CMMS data management. Maintenance and engineering management has come a long way over the last 30 years. Once regarded as strictly a reactive, resource-intensive cost center, maintenance organizations now run more like businesses. They collect and analyze data to assess, plan, and make important decisions aimed at preserving facilities and assets in the most efficient manner possible.

By Frank Lucas  

Work orders serve as the primary source of this information, due to their role in recording maintenance activities. But how do managers know that key facts and figures extracted from these documents are accurate and credible? What if the data on which managers base important decisions do not portray a true profile of the organization?

Computerized maintenance management systems (CMMS) have become invaluable tools in a manager’s arsenal, but they only produce output based on data that has been put into them. Just as important is the way technicians initially identify data for future recall. If handled incorrectly, this component can make even good, factual information tell an inaccurate story.

The heart of a maintenance organization lies in events that happen in the field. But problems arise when an organization doesn’t take the necessary steps to ensure its data-management strategies accurately describe these activities.

Managers must make certain that they thoroughly plan and execute this key process if they want to make decisions based on valid and reliable information. Despite advances made on the business side of facilities management, it still takes people to develop and execute procedures, conduct training, and perform review functions that reliably regulate data collection so it produces legitimate results.

Technicians’ Comments

Essential elements of a work order include its type, priority codes, staff hours, cost, and equipment tag numbers. But a solid case can be made that a technician’s field comments represent its most important component. As shown in the accompanying article, field comments communicate steps taken to complete the requested work.

These comments turn the rough draft of the work order into the final report, transforming forecast data into historical data.

Managers most often use this information to statistically profile the maintenance organization. Technicians also use this same data to review actions, uncover trends and determine root causes of problems.

So a technician’s field comments are more than just a historical record of steps taken to resolve work orders. They are the principle source from which a CMMS collects, identifies and manages statistical information.

In some instances, technicians already know a work order’s expected resolution, and they initially can code it to meet that expectation.

For example, a locksmith receiving a work order requesting a duplicate key probably will not resolve it in any way other than cutting a key. Chances are very good that no changes to the work order classification will be necessary if the technician coded it correctly to begin with.

But what about a work order to address a “too hot” complaint? Technician activities to resolve the complaint can range from simple adjustments or moderate repairs to a capital project to replace a piece of equipment.

Each of these actions describes a vastly different scenario for resolving this common service call. So at close-out, supervisors must review codes used to create a work order and modify them to appropriately reflect the work performed. Failing to do this triggers a data defect continuum of sorts, a chain reaction of events that can profoundly affect statistical profiles and eventual decision making.

Data-management Controls

To operate like successful businesses, maintenance and engineering organizations must have process and procedure documents that govern all aspects of data entry for work orders. These documents should include definitions for classifications and codes that management has deemed appropriate.

Along with these definitions, managers should include common examples of each activity to help employees understand the way that corrective actions taken dictate their use. For instance, a work order type called “repair” must include the organization’s definition of “repair.” It also should include examples of repairs that routinely occur in the facilities. Examples of traditional repair activities include replacing worn components, patching leaking pipes and re-wiring faulty circuits.

During the life of an active work order, at least two classification and coding opportunities occur — during initial creation and at close-out.

The creation of a work order offers the first opportunity based on known facts, which are usually minimal or general. In the case of the “too hot” complaint, a manager will not know the actual repair until the technician can assess the situation, take corrective action, and verify that he or she resolved the problem.

An organization should select default codes when defining factors that technicians do not know initially, with the understanding supervisors will review them at least one more time, at close-out. A technician initially might classify a “too hot” complaint as routine or recurring work, but a supervisor might change the classification at close-out after reviewing corrective actions documented in the work order.

Depending on the processes outlined in an organization’s control procedures, additional classification and coding opportunities are possible. Some organizations make interim changes to work orders as they pass through various stages of the work process. These can entail updating status codes, modifying job scopes, or routing line-item tasks to other shops and trades.

Such updates help managers control maintenance backlogs and schedules by making work order data as up to date as possible. Once a technician turns in a work order for completion and enters all postings and documentation, a final code review can take place. This step includes reviewing technician comments, ensuring he or she adequately describes the steps taken, and changing classifications and codes to more accurately match the corrective actions.

Supervisors then use classifications and codes to quantify data into various statistical categories that managers have determined will provide a representative snapshot of the department and will help them make appropriate decisions. This process assumes that technicians classified and coded properly and that supervisors reviewed the data for accuracy before it becomes part of the historical record.

Maintenance managers can select certain identifiers based on past experience, preferences, industry standards, and best practices, as well as those that fulfill site-specific requirements. Once managers select these identifiers, it is important that everyone understands the circumstances and situations that determine their use.

Examples of more common classifications and codes include request and work-order types, priority codes, building and room designators, statuses, trades and shops, equipment and asset tags, and condition-cause-action codes.

Other identifiers might stem from operational features built into the CMMS design or from requirements of other business systems that share data with the CMMS. These identifiers can include repair or cost-center labels, account-number formats, task or job numbers, warehouse and part identification, employee designators, timekeeping codes, and other site-specific references.

Frank Lucas is assistant director of work management for the University of Nevada in Las Vegas.

Next month: We walk managers through strategies for effectively analyzing data from work orders.

Data Difference: Forecast vs. Historic

Maintenance data breaks down into two main categories — forecast and historic.

A new work order actually serves as a rough draft that documents a maintenance concern or customer need. Managers then use this information to predict or forecast events and activities. Anyone who monitors maintenance backlogs or plans work schedules uses forecast data.

Conversely, a completed work order is the final report of steps taken to address the issue. This information becomes the actual or historic account of actions taken. Those who report on expended staff hours or accumulated costs use this historic data to generate these statistics.

In some cases, the same document preserves both forecast and historic data. An example is a project’s estimated cost, which is forecast data, and its actual cost, which is historic data.

In other cases, managers use forecast data to generate accurate historic data. For example, if a work order to answer a “too hot” complaint initially is classified as routine maintenance — forecast data — and a technician corrects the problem by making a repair — historic data — the technician also will need to modify the work classification to match that action.

Many maintenance managers fail to address this step in procedural documents and to execute it as part of daily operational routines. As a result, many reports and statistical compilations generated by the CMMS inaccurately represent events that take place in the organization.

Contact FacilitiesNet Editorial Staff »

  posted on 12/1/2005   Article Use Policy

Related Topics: