When a batch fails, you already know you’ll have to dig through both your equipment and materials to determine what went wrong. However, the order you check your records and the details you look at first can speed up, or slow down, how long it takes to get to a resolution.
Identifying the source of a quality issue is the foundation of any effective investigation. Before teams can assign quality corrective actions, address supplier performance concerns or prevent similar issues from recurring, they first need to determine whether the issue originated from equipment, materials or another contributing factor.
The five checks below can help teams quickly narrow down the likely source of a quality issue so they can investigate more effectively, take appropriate corrective action and reduce the risk of recurrence.
Cross-reference the lot across machines instead of the machine across lots
Check whether the defect's start time lines up with a machine event
Look at the actual measured value on the incoming lot instead of only pass or fail
Check the defect's physical pattern before pulling any records
If you’re still unsure, use recency as the tiebreaker
Here’s how to run each one.
1. Cross-reference the lot across machines instead of the machine across lots
This step is usually the fastest and it helps you start isolating whether it’s an asset issue versus a materials or supplier issue.
First, pull every batch the flagged lot touched. It doesn’t matter if the machine ran it or not. Pull them all. When a quality issue follows a supplier lot across multiple machines, manufacturers need more than traceability. They also need a structured process to investigate supplier performance, manage quality corrective actions and reduce the likelihood of future issues.
Your data hygiene is important here. This check works a lot faster when the lot numbers and machine IDs are captured and stored in the same record. If you store these records in different places, it’s going to take a lot more time and increase the risk that you miss information.
2. Check whether the defect's start time lines up with a machine event
Pull the maintenance logs for the machine in question. Once you’ve identified a likely cause, it's equally important to ensure the resulting investigation, quality corrective action and documentation are managed consistently. Look specifically at the window right before the defect started. (You can look at the full history if you want to, but you don’t need to.) If a tool change, restart, or completed preventive maintenance lines up with when the defect began, there’s a strong chance the equipment is the cause.
Similar to the first step, how quickly you can resolve this checkpoint depends on how precise your maintenance log timestamps are and if they’re easy enough to compare to your production timestamps. A log that only records "PM completed this week" instead of a specific time won't narrow anything down.
3. Look at the actual measured value on the incoming lot instead of only pass or fail
The next step is to pull the actual measured value for your lots. Go beyond looking at the pass or fail result, since a lot that technically passes incoming inspection doesn’t automatically eliminate it as the potential cause. Check where each lot sat inside the tolerance band. If a lot passed but sits close to the edge of your acceptable range, that’s a sign that lot could be the issue. Lots that pass more comfortably in the middle are less likely to be the culprit.
If the lot is confirmed as the source, look beyond a one-time fix. Determine whether supplier corrective actions, incoming material controls or inspection criteria need to be updated so the same issue doesn't reappear in a future lot.
We really emphasize not skipping this step. It’s an easy one to glance over since pass or fail is usually the easiest thing to check. Take the time to examine the actual measured values.
4. Check the defect's physical pattern before pulling any records
This step doesn't require pulling any records, and that’s what makes it fast. Lay out the rejected parts and look at where the defect shows up.
Does the defect show up in the same position on every part? That points toward a tooling or fixturing issue.
Does the defect show up on different spots on each part? That points more toward the material.
Identifying a recurring defect pattern is often the real starting point. Tracking that pattern consistently across investigations makes it easier for teams to spot trends, connect related issues and apply what they learn enterprise-wide, instead of treating every defect as a new, unconnected problem.
You can run this check faster than the first two checks on this list, but it can also be one of the least precise when used by itself. If you want, you can use this check to decide which of the other checks to run first, if you use it with other tests.
5. If you’re still unsure, use recency as the tiebreaker
If you run the first four checks and the results are mixed, then a nice way to break the tie is to go with the equipment or materials that you most recently changed. This could be a new tooling that you installed, a new lot you started or even a new operator running the job.
If everything was running fine the week before and now it isn’t after a recent change, then it’s probably not a sudden coincidence.
However, we would caution that you use this step as the one you fall back on and not the one you start with. After all, if you stop production and start with this step, then start back up again, you could miss something deeper happening within your operations.
Finding the source is the first step
Determining whether a quality issue started with assets or quality only gets a team partway there. Once the root cause is identified, organizations need a governed process to manage the investigation, assign corrective actions, verify they're actually effective and prevent the issue from recurring across other lines, sites or products.
This is where a dedicated quality management system (QMS) comes in. A QMS connects nonconformance tracking, corrective action management and enterprise-wide trend analysis into one governed process, so a finding from one investigation doesn't stay siloed in one team's notes. Organizations using Reliance for their QMS have cut quality CAPA resolution times by half, reduced scrap and rework by 32% and reduced warranty costs by 70%.
Next steps
Each of these five checks is simple to run on its own. The real work starts once you've found the source.
If you want to learn more about how Attune and Reliance can help you align manufacturing operations with your organization's goals check out [LINK: how Reliance extends the value of Attune to manage quality]. If you'd rather talk through your specific setup with an expert directly, we’re always here to help. Just schedule some time with us.