A tale of lost and found work order data
No data means working in the dark
The maintenance team at Voltalia had a big problem. They were closing over 104,000 work orders every year with no work order data to show for it.
The renewable energy producer and service provider had no idea if any of its PMs caused breakdowns instead of preventing them. Or if it was spending labor hours, parts, and other resources on unnecessary work. Or if it was assigning the right number of people to a task. Or much of anything about their work orders.
Rewriting the script to make data count
Voltalia’s maintenance team vowed to change this. After years of working toward their goal, they reached a “100% improvement in measuring maintenance KPIs,” in the words of Vasco Vieira, Voltalia’s Maintenance Engineering Director.
The data helped the company uncover some major efficiencies. For example, the work order data showed that one team was spending 40 hours every week driving from the office to an off-site facility. That meant adding time and costs to every job. The company ended up building a smaller satellite office near the off-site facility to save time and money.
The moral of the story
Work order data has the power to transform the way maintenance teams operate. There are the small wins, like making every job easier for technicians, that add up to bigger ones, like decreasing maintenance costs across the board. Voltalia is proof of that.
But this data is often overlooked. It’s not because maintenance teams think it’s useless. It’s because looking at thousands of work orders is not easy. This post provides some best practices for making this process easier so you can discover insights in your work orders and use them to make a difference at your organization.
How to get maintenance data from work orders
The most common obstacle to using work order data is having unreliable data or no data at all.
“Before you do anything with work order data, you need to know that it’s there and clean. If not, all the decisions you make afterwards are going to be flawed,” says Vishakha Shah, a Solutions Consultant on Fiix’s professional services team.
Getting off on the right foot with work order data is a four-step process:
#1: Define your goals
Some data is helpful. Too much is distracting. Having a goal will help you draw the line between the right numbers and the distracting ones. Some examples of a goal include:
- Building a world-class preventive maintenance program
- Creating a lean maintenance strategy
- Using maintenance to increase throughput across the business
#2: Create your measurements
Think about the areas of your day-to-day operation that can mark progress toward your goal. Some examples of measurements in your work orders include:
- Percentage of reactive vs. preventive work orders
- Number of faults found during PMs
- Frequency of reactive work on critical assets
- Number of expected vs actual labor hours
- Size of backlogged work orders
#3: Build work orders around those metrics
Set up your work orders to get the metrics you’ve chosen. To do that your work orders need to be created with three Ss in mind:
- Standard: Your work orders should ask for the same information every time. The process for creating, reviewing, assigning, prioritizing, and completing work orders should be as standard as possible.
- Specific: Be exact about what you want to know. For example, if labor hours are important, ask how much time each task took instead of the time for an entire work order. This will give you cleaner data and makes it easier to spot key metrics quickly.
- Simple: Involve staff who frequently make and complete work orders in the process. The input will help you design work orders that are easier to fill in and increases the likelihood they’ll actually be completed.
#4: Start small and scale your success
Finding problems in your process is heartbreaking when you’ve spent months on it. Avoid this by starting with work orders from one asset or from one area of the facility. Hone your measurements, get quick wins, and scale the process to other parts of the organization.
How to use work order data to find and fix problems
Collecting work order data is pointless if it’s not used to solve problems at your organization. Every facility has unique issues, but three most common ones are unplanned downtime, critical work that’s delayed, and work that takes more time and resources to complete.
How to prevent equipment downtime
Here are a few questions to ask to find the cause of reactive maintenance and how to make sure it doesn’t happen again:
- Was a follow-up task not created or completed? Make sure failed inspections trigger high-priority follow-up actions and alert the right people. A concise list of failure codes helps follow-up work be successful.
- Was a defective part used during a repair? Make sure other spares aren’t defective. If they are, follow up with your vendors to get new ones.
- Were tasks on a previous work order missed or done incorrectly? Review your task list and fix any unclear instructions that may have led to missed tasks. Supplement task lists by attaching asset histories, diagrams, pictures, and manuals.
- Was scheduled maintenance missed prior to the failure? Mark critical work as a priority and make it visible in whatever system you have until it’s done.
- Was production higher than normal/planned, was it done incorrectly, or was it modified?: Review your maintenance schedules and consult with the operations team to create stronger SOPs for when production increases or changes.
How to prevent work from being delayed
Work order data can help you find and fix work orders that took so long to get to:
- Parts and supplies were not available. Review the purchasing process for these parts, including minimum quantities and who can submit purchase orders so you’re never shorthanded again.
- The problem wasn’t identified properly or instructions were missing. See if the work order description, failure codes, and task list can be clearer. Attach photos, manuals, SOPs, or other documentation to the work order.
- An emergency work order diverted resources: This can’t always be avoided, but it could tell you that the task is too big. Consider breaking it into smaller tasks to prioritize parts of the job.
- There was a scheduling conflict with production: Talk to operations about why maintenance is necessary on the asset. Consider giving operators routine maintenance responsibilities associated with the work order.
Get operations on board with your maintenance strategy
- The person/people assigned to the work did not have the right skills: Make it very clear on the work request what kind of skills or certifications are necessary for certain maintenance types.
How to prevent work from taking longer than it should
Maintenance schedules don’t have a lot of room for error. When work runs long, it has a big domino effect. Work orders can give you insight into what’s causing work orders to take longer and how to fix the issue.
- It was assigned to the wrong person: Work will take longer if the technician didn’t have the right skill set. Standardized work requests let everyone know the right person to assign. Add as many manuals, pictures, diagrams, and other resources to work orders to help technicians who are unfamiliar with the task.
- The expected completion time was too low: The expected labor hours should be increased if a work order is consistently taking more time than is given.
- The task list was too big or unclear: Join an experienced technician as they complete the work order, document what they do step-by-step, and create task lists with this information. Give expected hours for each task in the work order so you know which ones are causing problems.
- Not enough technicians were assigned to this work order: It might not be a one (or two, or three) person job.
- Additional work was done during the work order: Develop processes that help technicians create and prioritize separate work orders for additional corrective repairs.
- Parts and supplies were hard to find: Bundle together all parts and supplies needed for common and critical work orders so they can be accessed quickly.
How to scale your success with work order data
What happens when you want to take the next step and start analyzing hundreds or thousands of work orders? There are three answers to this question:
- Prioritize work orders so you’re focusing on the ones that matter most
- Hire more people to analyze work orders
- Invest in a system that does all this for you
How to prioritize work orders
“If you’re strapped for time and resources, focus on reactive work orders,” says Stuart Fergusson, Fiix’s Solutions Engineer Leader.
Identifying how work orders contributed to failures will help you move toward a solid preventive maintenance program, says Stuart.
If you have reactive maintenance work orders locked down, the next batch to prioritize are high-risk, upcoming work orders. This is work that has the potential to go very wrong, including work on critical assets, work that hasn’t been done in a while, or large and complex projects.
If you can squeeze in a few more work orders, Stuart recommends analyzing work that costs a lot. Making these projects more efficient will make a major impact. Look at work that uses a lot of labor, major components, and planned downtime on production assets.
How to justify more resources for your team
Results are the currency you need to convince your boss that you need another person on your team. Highlight the problems you’ve uncovered and fixed by analyzing and optimizing work orders. For example, how many failures have you caught and prevented? Did you decrease the cost of projects by helping technicians be more efficient? No win is too small.
Show the impact of this success if it was achieved on a larger scale. If you saved a dozen labor hours on one work order, imagine how many labor hours would be saved across 100 work orders.
Drive the point home by describing the ripple effect this could have on maintenance. If someone could take work off your plate, it could mean less backlog. Or more training for operators to do routine maintenance tasks, freeing your team to do big projects. Focus on where a new hire may add value indirectly.
Software for work order analytics
Almost every maintenance analytics platform focuses completely on asset data. It’s not that easy to find a system that goes deep on work orders. Until now.
Work order insights, powered by Fiix Foresight, can analyze thousands of work orders in minutes and tell you what work has caused or will cause breakdowns, overdue work, extra labor hours, or other problems. The best part is, these insights are available on the Fiix analytics tool. That means you can see trends and predictions from a single dashboard without leaving your CMMS or messing around with spreadsheets.
Work order insights gives you recommendations by sorting through all your work orders, comparing similar ones, and identifying outliers to find the riskiest tasks.
For example, you might have many of the same asset, like a compressor, across multiple facilities. Those assets require hundreds of PMs per year. If the task count on one PM at one facility is half as big as all the others, work order insights will catch this. Once you’ve caught this problem, you can change your task list and avoid missing a crucial step in your scheduled maintenance.
That’s just a small taste of what work order insights can do. Learn more about how it works, what it looks like, and more here.
Everything you just read in three sentences
- Creating a successful work order data strategy includes defining your goals, choosing metrics and benchmarks that align with those goals, building work orders that collect those metrics, and piloting your approach.
- Studying work orders that took too long to complete or get to and were in response to breakdowns will help you identify the areas of the work order that need fine-tuning and prevent these problems from popping up again.
- Scaling your success with work order data relies on three things: Prioritizing work orders, quantifying success to justify more resources, and investing in work order systems that can take on tedious and time-consuming analysis.