RoxaneB
Thanks,could you share examples of criteria (triggers) to define if apply correction or corrective action? it sound very interesting
I think I could apply it for me, besides to use it to apply corrections/corrective action it may be useful as well to define when to raise NCR (nonconformity reports).
I start down the pathway of explaining triggers
here.
Examples and the triggers points must be unique to your organization (i.e., products, processes, culture, management system maturity, etc.) and based off of REAL data...no 'I feel' or 'I believe' or 'I think'.
Setting triggers also means that your organization is focusing their attention and resources on the issues that have higher risks/costs/impact/effect. In my previous job, we referred to a correction as a quick "wham, bam, let's get back to making steel" scenario. We fixed it, got back on track, and went back to work. Corrective actions, however, required more work with root cause analysis, action plans, etc. But either way, there was a nonconformance and it was recorded because each occurrence of a nonconformance is a data point to be part of a higher-level analysis.
You essentially identify as many nonconformance situations as possible and then, using real data, determine what the thresholds are for doing a correction, corrective action, and even preventive action, if you want. We documented this in a table/matrix, so that on a regular basis (e.g., annually as part of our annual and strategic planning process), we would re-evaluate and adjust our triggers accordingly. Sometimes, it was a balancing act between continual improvement and realism and priorities.
Current Year 2019
SCENARIO - After analysing your data, you decide to set some triggers. In the first example, let's say 80% of your complaints that want cash-back/rebates are valued at less than $100 and it will cost you more to investigate them than it would to simply pay the customer. In the second example, let's say you are noticing Machine A has a lot of unscheduled downtime because it's not set-up properly and in a timely fashion.
1. Simple and Generic Example
Customer complaint where they are asking for a rebate/cash-back:
- Correction - Rebate/Cash-back < $100
- Corrective Action - Rebate/Cash-back >= $100
- Preventive Action - Rebate/Cash-back from same customer > 5 times in a fiscal year
2. Less Simple and Generic Example
Machine A unscheduled downtime due to set-up delays:
- Correction - Delays < 30 minutes
- Corrective Action - Delays >= 30 minutes
- Preventive Action - >5 consecutive unscheduled delays <30 minutes within past 7 days (rolling calendar)
Next Year 2020
SCENARIO - Now it's time to start planning for the next year. You pull the data and notice that cash-back $$$ haven't changed, but that Customer XYZ increased from 3 complaints in 2018 to 50 in 2019, sometimes they want cash-back, but over 75% of the time it's because of poor packaging. However, their volume has also increased making them one of your more important customers. You do the math and see that Customer XYZ has moved from 1 complaint per 1,000 widgets shipped to 1 complaint per 100 widgets shipped!
As part of the analysis, as well, it is shown that Machine A's unscheduled down time due to set-up delays has improved quite significantly. 80% of the unscheduled downtime due to set-up delays is now < 20 minutes.
1. Simple and Generic Example
Customer complaint where they are asking for a rebate/cash-back:
- Correction - Rebate/Cash-back < $100
- Corrective Action - Rebate/Cash-back >= $100 OR Any packaging-related complaint from Customer XYZ (regardless of cost)
- Preventive Action - Rebate/Cash-back from same customer > 5 times in a fiscal year
2. Less Simple and Generic Example
Machine A unscheduled downtime due to set-up delays:
- Correction - Delays < 20 minutes
- Corrective Action - Delays >= 20 minutes
- Preventive Action - >5 consecutive unscheduled delays <30 minutes within past 7 days (rolling calendar)
@>--\---/----
Improvement on trigger thresholds will not magically improve. As part of our annual and strategic planning, we tied our projects, initiatives, and objectives to metrics which were often aligned with certain trigger metrics. Using the Machine A example, we would, for example, have a documented plan to reduce those set-up delays causing unscheduled downtime.
If you're not sure where to start with your list of triggers, look at your companies objectives. Hopefully, there are metrics that will be able to support the claim of meeting (or not meeting) objectives. Take those metrics and, using data, calculate your trigger thresholds.
