What are the consequences of GOOD data? If we can confidently say we have good data quality, our reports are spot on, our dashboards are telling us what we need to know, and we are making great decisions based on this perfect data. The result of good data is better, faster decision-making that increases the effectiveness of our maintenance programs and efficiency of our assets. All of which have a positive impact on the bottom line.
So what are the consequences of bad data? Typically the term ‘bad data’ refers to data that is inaccurate, incomplete and can’t be confidently used to make decisions. The big problem here is visibility into if your data is good or needs improvements. Many organizations don’t know their data is of bad quality and continue to use it to make reports, dashboards, and decisions.
We all want to get out of the “bad data” game and move towards what I’d like to call “digital fidelity”.
Digital fidelity is a concept of standards around the quality of the digital data we use to make informed decisions.
Better data will make positive impacts across your maintenance program and business. Here are a few of the impacts you will experience what you improve your data:
- Improved and more accurate reports (data quality)
- Automated reports that do not require a complete review of the data (data confidence)
- More accurate dashboards
- Our collective time is more efficient across the board
- Our decisions are based on facts more than gut feelings
- Our actions are more aligned with our objectives
- Our assets are more efficient
- Our programs are more effective
How do we improve the data we are using and ensure it is driving better decision-making within our maintenance program and overall operations? Here are six steps:
6 Steps to Improve Data Quality
1. Step One: Conduct a Data Assessment
The first part of your data improvement should be an assessment of your current data. Looking at your data in this way should be an ongoing initiative in your organization – identifying gaps in data entry, recording, processes, etc.
Elements to assess in your data assessment:
- What data is collected
- How the data is collected
- How the data is entered
- Data completeness – does the data tell the full story?
- Data validity – is the entered data following data guidelines such as format and range?
- Timeliness – when is the data collected?
You can even start by targeting a specific area where you feel you can have the biggest impact in the shortest amount of time to show a quick win.
Planning and scheduling and asset criticality are good examples of these quick win areas.
If you are in the process of upgrading or changing your EAMS/CMMS, I would strongly recommend you include time to do a full assessment of your data as part of the bigger effort. Your trusted EAM advisor should be able to walk you through this. You can also learn more about our assessment offerings here.
2. Step Two: Determine Well Defined Objectives around your Maintenance Data
You must have well-defined objectives that are aligned to the overall corporate objectives and site/department objectives.
Data objectives include:
- What is collected
- Why it is collected
- What is done with the data
If there is no alignment, data quality, accuracy and completeness will suffer.
You may end up wasting time collecting data that is not needed or you might miss critical data that will take additional time to gather. Make sure all team members are on the same page with the objectives and that they are aligned with the maintenance processes.
3. Step Three: Define and Document What Data You Need
Define what data elements you want to track for each type of record you have, i.e. assets, spare parts, work orders, etc. If we take assets as an example, document what data you want to capture for assets in general and specific data for each class of asset and what type of standards and attributes you will put in place. This gets tricky, but most companies have asset classes identified with a list of attributes for each class already in place, just not being used. If you have criticality on all assets, is it well defined and based on a matrix that makes sense? Based on your industry, similar assets can have very different criticality rankings. Criticality on assets can definitely help identify what work to plan and schedule first and what spare parts should be more readily available than others (note that planning and scheduling are two different things).
Make sure that your processes are documented and that your data definition is also well defined and documented. Value lists are great ways to ensure the accuracy of the data and should be used as often as possible. If you do a good job with the assessment and the objectives, you will know what data you are missing and you will understand what you need which takes us to the next step. These fields should be included and in some cases required upon entry.
4. Step Four: Determine How You Will Gather/Collect the Data
This sounds simple, but based on what technologies you have in place, it is very important to document how data gets into your systems. When I was in college, we had a paper-based work order system with all the instructions on what data to print (with a black ballpoint pen and in block letters) next to each field. It was a very simple system that was well defined. As for results, it took six months to get a response to the only work order I ever submitted. I wish I had a picture of it. The gist was that my desk light was not secure and did not work. The response – tell the light to get a job and seek counseling!
Over the years, I have been involved in many data enhancement efforts and in each case, it was imperative to define how the data will be gathered. Specific types of data can be loaded into your EAMS/CMMS manually or automatically via an external integration. Make sure that if any of these processes change, the documentation is kept up to date. You may also require technicians to enter data manually, via handhelds, or by writing comments on the work order. One of the best ways to ensure that data is entered properly and efficiently is with handheld technology that will populate once online.
Another tip here is to reduce human error by choosing from option sets versus blank data entry whenever possible. Predefined option sets reduce spelling errors that largely impact data quality.
5. Step Five: Define How You Will Maintain the Data
If you have a dedicated data department focused on maintaining the data, select a data steward to manage data quality. There should be standards and processes around how data is managed on a regular cadence. Establish dashboards, metrics, and KPIs that are aligned with your processes and objectives. These should show the status of your overall data and allow you to drill down into the data sets that make up the data, allowing you to identify gaps and areas for improvement.
Remember, we are using (or should be using) the mass amounts of data generated from our assets, maintenance, and operations to help make better decisions for our business. Its imperative that the data be continuously monitored and managed for business growth and continuous improvement.
6. Create a Culture of Continuous Improvement
The effort to improve data and make better business decisions should be an ongoing practice as part of your organization’s continuous improvement initiatives.
Once your data is set and you have proper views – metrics, KPIs, and dashboards – it becomes easier to look for new areas to improve. For example, if you have job plans in place with the appropriate data, it becomes easier to evaluate the performance of those job plans and adjust them if there is a variation between planned hours and actual hours. It also becomes easier to implement PM Optimization without any false starts due to bad data.
The Impact of Having Better Data in your Maintenance Program:
What are some benefits we can expect to see? The following is a high-level list that comes to mind. There are many others that you may experience based on your specific objectives.
- Improve Asset Utilization and Performance
- Reduce Capital Costs
- Reduce Asset-related Operating Costs
- Increase Uptime
- Reduce Failures
- Improved efficiency – assets and resources
- Improved Safety
- Extend Asset Life
- Maximize Overall Asset Productivity
- Minimize Total Cost of Ownership
- Know whether it is more cost-effective to continue to maintain, overhaul or replace a failing asset.
The bottom line: Data Quality is worth the effort!
Note: This article was previously published on September 21, 2021, on our sister company’s website.