30-day RFID pilot
Dispatch door 1
- Read point
- Dispatch door 1
- Tagged
- Shipper carton
- Tag
- Paper UHF label, ₹6, the one you would buy in volume
- Dates
- Four weeks from the start date
- Owner
- Hand count
Pass marks
| Measure | Pass mark | What it takes |
|---|---|---|
| Read rate per tag | 99.90% or better | At most 1 missed in about 4,800 passes over weeks 2 and 3. |
| Loads read complete | 95% or better | About 96 loads of 50. At 99.9% a tag, expect 95.1% complete. |
| Stray reads that become events | 0 | Every raw stray read is logged with its time and the door it really used. |
| Tagging time | 5 s an item or less | Timed in week 1, by the people who'll do it. |
| Traffic | No change | Same forklifts, same speed, same route. |
Week 1: tag and baseline
- Fix the one read point. Everything else in the plant stays as it is.
- Buy the tag you would buy in volume, from the supplier you would use. No vendor samples.
- Tag real stock on the real line, at the position the tag will have in production, by the people who will do it.
- Time the tagging. Seconds per item, per person.
- Count by hand what passes the read point, every shift, against the ERP. That count is the baseline.
- Write the pass marks below on this sheet and sign it before any reads come in.
Weeks 2 and 3: run it beside what you have
- Switch nothing off. The manual count and the ERP carry on as before.
- Every pass: what the reader saw, what the count says, what the ERP says.
- Log every disagreement: which item, which load, which shift, where the tag sat, what was next to it.
- Log every read of a tag that did not go through this point. Those are stray reads.
- Same forklifts, same speed, same drivers. If the reader needs them to slow down, that is a finding.
- Change nothing mid-run. If something has to change, note the time and start the count again.
Week 4: the sums
- Read rate per tag: tags read over tags that passed.
- Loads read complete: loads with every item read over all loads.
- Stray reads, and how many of them became an event the ERP saw.
- Where the misses cluster: middle of the load, against metal, under a tine, one shift.
- Compare each number with the pass mark written in week 1. Not with what the vendor said.
If a mark is missed
Change one thing, then run weeks 2 and 3 again. Two changes at once and you won't know which one worked.
| The log shows | Change |
|---|---|
| Misses cluster in the middle of the load | Tag position or a tag with more sensitivity. Then antenna angle. |
| Misses on items against metal or liquid | An on-metal tag (about ₹60) for those items, or move the tag. |
| Misses on one shift | How the load is built or driven on that shift. Not the radio. |
| Stray reads becoming events | The rule that turns reads into events: a second antenna for direction and a beam break at the door. |
| Tagging is too slow | Pre-printed and encoded labels at the supplier, or a printer-encoder at the line. |
Pass marks agreed before week 1.
Signed
Date
Made with the RFID pilot planner, iotready.com/tools/rfid-pilot-planner