AI agent for quality control inspectors
Gauge Repeatability Study Agent
Prove that each gauge is reliable before its readings are trusted.
What it does
A gauge can look fine and still give readings that vary too much to trust. This agent plans a measurement study for each gauge: how many parts, operators and trials. It collects the measurements, calculates repeatability and reproducibility, and flags gauges where the variation is too large for the tolerance. If a gauge fails, it asks for calibration, repeats the study afterward, and rechecks before the gauge is used. The inspector approves the actions. It keeps a history of each gauge, so a slow drift shows up before it fails. A gauge that fails is not used until the repeat study passes. Edge case: a gauge that passes overall but shows one operator reading differently is flagged for training.
How it works
Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.
Read the steps as a list
- Study is due
- Plan parts, operators and trials
- Collect the measurements
- Calculate repeatability and reproducibility
- Is variation within the accepted share of tolerance?If not: Flag the gauge and ask for calibration or repair. Back to step 2.
- Request calibration and wait for the record
- Repeat the study after calibration
- Does the gauge pass the repeat study?If not: Take the gauge out of use and ask for replacement. Back to step 7.
- Check operator differences
- Inspector approves the actionsThe agent waits here for your OK.
- Gauge study report
How it decides
It passes a gauge when its variation is within the accepted share of the tolerance, and requires a repeat study after any fix.
- Accept below 10 percent of tolerance
- Review between 10 and 30 percent
- Reject above 30 percent
- Always repeat after calibration
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Accept and reject limits (default 10 and 30 percent)
- Study size
- Study frequency
- Gauge list
What keeps you in control
It always asks you first
- Inspector approves the actions on each gauge
Hard limits
- Never use a failed gauge
- Never skip a repeat after calibration
It stops when
- Done: gauges are passed or out of use
- Stop: the study cannot be run and the inspector decides
Set it up
We guide you through the set-up, step by step
Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.
- One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
- The agent then walks you through connecting your own data, one source at a time
- A downloadable copy with the flow chart, the rules and the full guide