Comparison, correct as of August 2026
RFID against computer vision for stocktaking.
RFID reads a tag without needing to see it, which works inside a case or through packaging. Sentispec Inventory reads the label already on the pallet. The choice turns on whether you need item level tracking or pallet level location accuracy, and on which of the two costs more as you store more.
The case for RFID
RFID reads what a camera cannot see.
RFID reads a tag it cannot see, at speed, through packaging. That matters wherever the requirement is finer grained than a pallet.
- No line of sight required. A tag inside a shrink wrapped pallet, buried in a case, or facing the wrong way on a shelf can still be read by an antenna. A camera needs the label visible and legible.
- Item level granularity. RFID can identify individual units inside a case, not just the pallet or the location, which suits high value or highly regulated items where every unit needs its own record.
- Portal reads at speed. A fixed reader at a dock door or gate can read every tagged pallet or case passing through automatically, with no separate scanning pass and no vehicle needed at all.
Side by side
What each method costs to keep running.
| Decision axis | RFID | Sentispec Inventory |
|---|---|---|
| What identifies the item | A tag read by an antenna | The label already on the pallet, read by a camera |
| Tagging labour required | Every pallet, case or item needs a tag applied and maintained | None, the unit is taught the labels you already use |
| Line of sight | Not required, reads through packaging | Required, the label must be visible |
| Reliability near metal and liquid | Degraded, metal and liquid absorb and reflect RF signals | Unaffected, it is a visual read |
| Untagged stock arriving | Invisible until someone applies a tag | Read immediately, no preparation needed |
| Best granularity | Item level, inside a case | Pallet and individual racking location, including empty ones |
| Capital equipment | Tags, printers, handheld or portal readers | None, the camera is leased inside the annual fee |
| Deployment | Vendor and tag rollout dependent | Live in 5 days |
Where RFID is the better choice. If the requirement is item level, if products travel where a label cannot face the camera, or if the value comes from a portal read at a doorway rather than a pass down the aisle, RFID is the right tool and the tagging cost is worth paying.
Where the camera wins
Nothing has to be tagged before it can be counted.
No tagging labour
Stock that arrives from a supplier without a tag is invisible to an RFID system until someone applies one. A camera reads the label that is already there.
No metal or liquid problem
Dense racking, shrink wrap and liquid filled products all interfere with an RF signal. A visual read is unaffected by any of them.
Location level, empty positions included
Location level recognition reports empty positions as well as occupied ones, which a tag based system only sees if the tag itself is missing.
The cost model against counting by hand is on manual stocktaking, and the same camera against a drone or a ground robot is on inventory drones, warehouse robots and why Sentispec.
Questions we get asked
RFID against computer vision.
Can we use RFID and Sentispec Inventory together?
Yes, they answer different questions. RFID can hold item level detail inside a case, while Sentispec Inventory holds pallet and location level accuracy across the whole racking, reconciled against your WMS with the recording behind every line.
What happens to stock that arrives without a tag?
With RFID it stays invisible to the system until someone applies a tag by hand. With Sentispec Inventory it is read on the next pass, because the camera works from the label the pallet already carries, with zero integration effort to start.
Does dense racking or liquid stock affect either method?
Metal racking and liquid filled products can absorb or reflect an RF signal and reduce read reliability. A visual read is not affected by either, though the label still has to be visible to the camera.
Next step
Tell us what granularity you need.
A 30 minute demo on recordings from a site with your racking and your label format, and a straight answer on whether pallet level or item level is the right target.