ChatGPT

Yep — **you *can* use your own ratings**, but there are some important conditions to get right so you don’t get burned in audits.

Here’s the clean breakdown for **AIAG & VDA DFMEA** 👇

---

## Short answer (but done right)

**Yes, organizations may use their own Severity, Occurrence, and Detection rating scales** *instead of* the AIAG & VDA grids **if** they are:

* Clearly **defined**
* **Consistently applied**
* **Documented and justified**
* **Equivalent in intent and rigor** to the AIAG & VDA approach

---

## What AIAG & VDA actually require

The AIAG & VDA FMEA Handbook **does not mandate** that you must use their exact numeric tables. What it *does* require is:

* A **structured, systematic rating system**
* Ratings that reflect **real engineering risk**
* Alignment with **customer and regulatory expectations**
* A method that supports **prioritization and action** (Action Priority logic)

So the handbook gives **recommended grids**, not untouchable law.

---

## Conditions you MUST meet if you customize ratings

### 1. Severity is the most sensitive

Severity should:

* Still be **independent of controls**
* Reflect **customer impact, safety, and compliance**
* Be aligned with **customer-specific requirements**

⚠️ Many OEMs **do NOT allow customization of Severity**.
Always check CSRs (Customer-Specific Requirements).

---

### 2. Occurrence & Detection are more flexible

These are most commonly customized and usually acceptable when:

* Based on **your actual process capability, historical data, or validation methods**
* Scales are **clearly mapped** (e.g., what makes a “3” vs a “6”)
* Detection reflects **design controls**, not manufacturing controls (for DFMEA)

---

### 3. Action Priority must still work

Even with custom ratings:

* The **AP logic (High / Medium / Low)** must still function logically
* High severity + poor detection must still drive action
* No “gaming” the numbers to reduce APs 😄 (auditors hate that)

---

### 4. Documentation is non-negotiable

You should have:

* A **company FMEA manual or procedure**
* Your **custom rating tables**
* A statement explaining **why** they differ from AIAG & VDA
* Evidence of **training** on the custom system

---

## What auditors usually accept

✔ Custom scales used consistently
✔ Clear definitions tied to engineering reality
✔ Evidence they improve risk understanding

## What auditors usually reject

✖ “We just changed the numbers”
✖ Ratings with vague wording
✖ Scales that conveniently lower Action Priority
✖ No written justification

---

## Best-practice recommendation

Many companies do this:

* **Use AIAG & VDA Severity as-is**
* Customize **Occurrence and Detection**
* Include a **mapping table** showing equivalence to AIAG & VDA

That keeps customers, auditors, and internal teams happy.

---

If you want, tell me:

* Your **industry** (automotive OEM, Tier 1, medical, aerospace, etc.)
* Whether you’re dealing with **specific OEM CSRs**
* Whether this is for **audit defense** or **internal improvement**

I can help you sanity-check your custom rating grid or draft a justification that auditors won’t argue with.