One of the most important functions of a CMMS application is the ability to make sense of the reams of data collected by the software and its many points of integration. This functionality goes by many names, from simple report and graphics generation; to the more sophisticated decision support systems (DSS) and executive information systems (EIS); to data mining, data marts, and data warehouses; and so on.
We have seen quite a progression in how CMMS vendors allow users to extract data in a meaningful way. Regardless of which advanced tool is deployed, the key is to analyze and present data so that effective decisions can be made automatically (e.g., a control loop) or by a given function or individual. Of course, an important assumption is that source data input is accurate and timely—this may or may not be problematic in your environment.
Popular buzzwords like data analytics, artificial intelligence, business intelligence, dashboards, and scorecards together encompass a continuously improving and powerful tool for managing the massive amount of data collected by CMMS software today. In some cases, CMMS vendors have opted to build their own engine to provide the functionality. This affords seamless integration with the rest of the package but may not be as sophisticated. Others encourage the user to buy specialized and sophisticated add-on software from vendors; however, this requires purchasing and learning another application. Finally, some vendors supply these add-ons through third-party license, which may be the best of both worlds if the package is properly integrated.
Why the hype? Think of the dashboard of your car, and then try to imagine what would be on a “dashboard” for your CMMS application. Imagine an online, real-time summary of the latest scorecard results including various ratios, forecasts, trend graphs, and estimated versus actual cost comparisons. Picture the use of color to denote if trending is out of range; for example, an indicator remains green if a given measure is OK, yellow if a measure is beginning to trend outside of the acceptable range, and red if it is out of range. Measures might be levels of downtime on a given piece of equipment or number of PMs past due, or the level of spare parts inventory. In terms of graphics, indicators may be in the form of speedometers, lights, dials, charts, graphs, and so on.
Once the user is presented with a yellow or red “condition,” there is an option with some vendors to drill down on the indicator for more detailed reporting on the source of the problem. Action is then required to bring the measure back in line.
Some CMMS vendors have developed a business intelligence system that encompasses a whole hierarchy of scorecard measures and related dashboard graphics, to allow you to see your maintenance operations at a glance. For example, suppose at 11:00 am, you notice your highest-level, overall indicator pushing 7.5 out of 10, which causes the indicator to turn yellow and start to flash—and your goal was to maintain an overall score of 8.5 this year. You can then drill down to sub-indicators under the headings, say, asset performance, production, inventory, human resources, and financial. Of these, perhaps all indicators are green except for asset performance, which shows say 4.3 out of 10 and is clearly condition “red.”
Successive drill-downs may eventually indicate that there is a problem with the reliability of one of the assets in that a problem code has re-occurred at great cost for the fifth time on several of the same production lines. As well, there has been a serious safety problem resulting in a lost-time accident. These lower-level indicators were both condition red, and thereby causing the higher-level variances. It is then possible to drill-down still further and access base documents such as work orders and cost summaries.
This balanced scorecard approach to monitoring your maintenance operations is unquestionably an extremely effective management tool. However, in my view it requires a relatively sophisticated management team and user base to set up the hierarchy, measures, control limits, and graphics relevant to each user group. Furthermore, and perhaps more importantly, data must be collected accurately on a timely basis, and results must be understood and followed up by all.
Although significantly advanced analysis and reporting tools have been developed by CMMS vendors in the past few years, this is clearly an area where the packages have tremendous opportunity for still further improvement. It is not fair to solely blame the vendors for the gap, as users are not yet demanding sophistication in this area. This stems from the users' lack of readiness for computerization as evidenced by the still-high failure rate for CMMS implementation. Moreover, the average utilization of features and functions of a typical CMMS runs at about 15-30%, and this percentage is only getting smaller as CMMS packages get more complex.
Thus, to make the business intelligence, dashboard, and scorecard features more effective, they must provide a simple tool that helps answer the key question, “What’s in it for me?” for each user. The CMMS must also support management in making decisions on what to do about a given data output. The latter is where artificial intelligence is just beginning to support the diagnostics, analysis, and decision-making regarding that data output.
In answer to the former question, the following provides a sampling of some of the many features and functions that could provide the balance between sophistication and simplicity:
Work order control. The work order is the focal point of any CMMS. It is therefore critical for users to have access to as much information as possible when entering work order data. Several systems provide access to parts on-hand, on order, on reserve, in transit, in repair, and in QA inspection. As well, a high-end CMMS provides analysis of tradesperson utilization, work order history, work order and part status statistics, asset performance, and even a troubleshooting and database and diagnostic assistant right at the point of data entry. Business intelligence can be used to assist maintenance planners to more easily determine the appropriate action to be taken and timing required.
Preventive and predictive maintenance. This is the most important area for many maintenance shops. Some of the more sophisticated features offered are multiple PM triggers, schedule flexibility (e.g., seasonality, multiple formats, zoom, and simulation), and condition monitoring for user-defined data (e.g., activating a PM work order when meter readings reach a certain value). Business intelligence can be very effective in monitoring output from preventive and predictive maintenance programs to ensure measures are trending within acceptable limits, and identifying what work must be done when.
Materials management. CMMS vendors vary widely in terms of the sophistication of this function. Some of the more advanced features include multiple costing methods, ABC and XYZ analysis for classifying your inventory, multi-warehouse tracking, multiple part number cross-references, serialized component tracking and integration with e-procurement. Business intelligence can provide an analysis of inventory and supplier history including what-if analysis on service levels, thereby allowing users to fine-tune the balance between service (i.e., stockouts) and cost (i.e., inventory levels).
Asset management. A successful CMMS implementation produces savings and benefits that stem ultimately from proper asset management. Equipment history reports on actual versus planned labor, material and other costs. Business intelligence can provide more advanced features that include tracking maintenance costs by user-defined statistics (e.g., cost per volume produced), equipment status tracking and analysis, and complaint, cause, action and delay code analysis. Other important features are analysis of production versus machine downtime, mean-time-between-failure (MTBF), drill-down capability to determine the cause of downtime on summary reports, and analysis of total cost of ownership in support of repair/replace decisions.
