Define each step in an event
Keep the reader response, parsed identifier, check-in decision, saved event, and acknowledgement separate. A repeated reader report should not automatically create a second check-in.
Entry 07 / Embedded hardware
Stockroom RFID check-in prototype.
A read is an observation. A check-in is a decision.
Planned · not startedPROPOSED ARCHITECTURE / RFID
- 01READERRaw observation
- 02PARSEIdentify item
- 03DECIDEAccept / reject
- 04STORELog & confirm
DEFINE → IMPLEMENT → VERIFY → DOCUMENT
01 / The work
A read is an observation. A check-in is a decision.
At USPA, I work with RFID inventory workflows and help isolate device, application, and data problems. This independent project would let me explore those interfaces in a system I can document openly. It has not been commissioned or deployed by USPA.
Select the reader before designing the interface
The selected module will determine supply requirements, logic levels, connections, and command framing. This project uses an external reader; it does not claim the design of an RF front end or antenna.
Choose the identifier deliberately
EPC identifies the associated object. TID describes the tag chip and may include a chip serial number. Document which identifier the prototype uses and how a synthetic item record is associated with it.
Make recovery visible
Use status indicators and an event log to distinguish received, accepted, saved, and failed states. Test interrupted communication and storage before drawing conclusions about reliability.
Proposed validation / no tests run yet
How I plan to test it.
- 01
Exercise valid responses, duplicates, truncation, malformed fields, timeouts, and reconnection with a mock reader.
- 02
Verify one accepted event per item and check-in session under a written duplicate policy.
- 03
Interrupt power or communication around storage and check whether saved and unsaved events are accurately distinguished.
- 04
After assembly, verify supply behavior and serial communication with the selected reader.
- 05
Report physical trials with distance, orientation, tag-group conditions, and counts of reads and misses.
Technical references
GS1: EPC and tag-identification questions