Test principle
Define the read zone, expected population and business decision first. Then test representative tags across materials, orientations, operator paths and exception conditions. Report captured, missed, unintended and duplicate observations separately.
Define the pilot question and ground truth
Write one decision for each scenario, such as whether a cycle count can identify the expected shelf population without unacceptable adjacent-aisle reads. Count and identify the expected tags before each run.
Record reader model, regional variant, firmware, SDK, antenna, power, session, filters, host and application build. A result without configuration cannot support a purchase decision.
- Scenario and decision owner
- Expected EPC list or controlled count
- Reader and software baseline
- Acceptance criteria before testing
Build the tag and environment matrix
Cover tag inlay and encoding, item material, packaging, orientation, stacking density, height, distance and nearby metal or liquid. Include representative difficult and normal populations.
Do not change several variables in one run. Use repeated runs and rotate operators or run order when operator technique may influence the result.
- Tag and attachment method
- Material and package geometry
- Orientation and density
- RF interference and adjacent zones
Execute a repeatable operator route
Mark start position, sweep path, distance, duration and stop rule. Train the operator to the documented method before recorded runs.
Capture raw events where permitted, then derive unique expected tags, missed expected tags, unintended tags, time to first required capture and total task time. Keep customer data out of ordinary logs.
- Fixed route and session duration
- Multiple repetitions
- Expected versus observed comparison
- Exception and recovery notes
Add battery, thermal and recovery evidence
Run long enough to represent the intended shift model. Record reader and host battery separately, temperature or throttling observations, disconnects, sleep recovery and offline synchronization.
Collect operator feedback on balance, trigger, screen visibility and fatigue, but keep subjective feedback distinct from RFID capture measurements.
- Reader and host energy use
- Thermal or performance change
- Sleep and disconnect recovery
- Ergonomic observations by operator
Pilot result worksheet
Report each metric by scenario and repetition instead of one blended success number.
| Metric | Definition | Why it matters |
|---|---|---|
| Expected unique captured | Expected tag IDs observed at least once | Measures coverage against ground truth |
| Missed expected | Expected tag IDs never observed | Shows workflow exceptions |
| Unintended unique | Observed IDs outside the defined zone | Shows over-read risk |
| Time to required set | Time until the scenario completion rule | Connects RF behavior to task time |
| Recovery events | Disconnect, sleep, retry or app recovery | Exposes operational fragility |
UHF RFID pilot checklist
Do not publish a performance claim unless the complete method and limitations can travel with it.
- 01
Business decision and read zone defined
- 02
Ground-truth tag population verified
- 03
Regional frequency configuration confirmed
- 04
Reader, host, OS, firmware and SDK recorded
- 05
Tag, material and orientation matrix prepared
- 06
Power, session and filter settings fixed
- 07
Operator path and stop rule documented
- 08
Repeated runs completed
- 09
Missed and unintended reads reported separately
- 10
Battery, thermal and recovery observed
- 11
Limitations and acceptance decision reviewed
Frequently asked questions
What is a good UHF RFID read rate?
A useful threshold is project-specific. It must reflect the expected population, permitted misses, unintended reads, task time and exception process.
How many repetitions are required?
Choose repetitions based on variability and decision risk. State the number in advance and report every run rather than only the best result.
Can a manufacturer range claim replace a pilot?
No. Listed range is useful for screening, but tags, materials, regional power, orientation, environment and workflow determine useful project performance.
Evidence and limitations
RuggedLayer test method / no measured result claimed
This method is a planning framework. Any future result must name device, configuration, method, repetitions, date, author, reviewer and limitations.
Reviewed: 2026-08-10
- RuggedLayer Test Lab evidence standard
- Manufacturer-supported DT50P Lite and RFG91 model records verified 2026-08-02
