"With RFID, you just wave a reader around and the stocktake is done, right?" That's many factories' first impression. In reality the reader is the last step. What decides success is the planning before it: how tags are attached, how IDs are structured, and how the data lines up with your systems. In 2023 we worked with the Stone & Resource Industry R&D Center (SRDC) to use UHF RFID on outdoor raw stone and upright slabs at a stone plant. Here's what we learned, turned into seven things any factory can settle before starting.

1. First: do you need RFID, or are barcodes enough?

RFID isn't automatically better. It solves the specific things barcodes can't:

Barcode / QR codeUHF RFID
How it's readMust be visible and aimed at, one at a timeNo aiming needed; many tags can be read at once from a distance
When items are blocked or stackedCan't scan what you can't seeOften still readable (depends on materials and environment)
CostA printed labelExtra cost for tags and readers
Best forSmaller quantities of items you can hold and scanLarge quantities of heavy, stacked or hard-to-reach items

Stone is a textbook case. Raw blocks weigh over ten tonnes and are stacked outdoors; slabs stand side by side, only 2–3 cm thick, so tags can only go on the edge. With barcodes, a stocktake often means climbing up or pulling slabs out one by one — exactly where RFID adds clear value.

2. How far away must tags be readable? That decides which RFID

RFID comes in different frequency bands; the ones you'll hear about most are UHF and NFC (HF). Roughly:

Write the requirement down as a number. At the stone plant, ours was "read raw stone and slabs from at least 4 metres away", so we chose UHF tags with handheld readers. With a number, you can discuss it with tag suppliers and verify it on site.

3. Define the ID format before choosing tags

This step is the most often overlooked and the hardest to change later. Every tag chip has a factory serial number (TID), and many systems use it as the item ID. We deliberately didn't. Instead, the barcode number is the primary ID, written into the RFID tag and printed on the label.

That makes tags reusable. Stone goes from raw block to slab to finished product, so items keep changing. If the ID were tied to the chip, each tag would be used once and thrown away. With the ID following the barcode, a tag's data is cleared at the next stage and the tag is reused — without duplicating inventory or transaction records.

SRDC also coordinated an industry-wide barcode format: 17 characters for raw stone and 20 for slabs, covering company code, material code, year-month and serial numbers. An ID scheme is a language the whole plant — or industry — shares, so it's worth getting right first.

4. Design tag mounting separately for each environment

In one plant, different items may need completely different mounting. For stone:

We judged mounting methods on three criteria: it sticks firmly, it lasts, and staff can do it quickly. The third is often forgotten, but if a tag is too fiddly to attach, it won't happen on the floor. Don't expect one tag to fit everything.

5. Use RFID and barcodes as backups for each other

Tags fall off or break, and readers run out of battery. Each slab carries both an RFID tag and a barcode; each raw block gets two tags with the same barcode, so one can fail. Printing the barcode on the RFID tag also lets staff confirm the number by eye, with no device at all.

6. Readers: handheld or fixed?

Most factories start with handhelds, then add fixed readers at key entrances once the process is stable.

7. After the read: system integration is what matters

RFID only tells you which tags were read. For a stocktake to be useful, that has to be matched against inventory data. Our flow:

  1. Download the inventory for one storage location to the handheld reader.
  2. Read the tags at that location; the app produces a stocktake list of what's there and what's missing.
  3. Tap an item in the list to see its data and photos.
  4. While stocktaking, upload the latest location back to the ERP, so location data never goes stale after items are moved.

RFID can be deployed on its own or integrated with an existing or new ERP. Either way, it needs clear storage-location numbering first; see storage-location visualization in stone plants.

Finally: pilot on a small scale

Real-world RFID performance depends heavily on materials, environment and how items are placed; the read range on a spec sheet isn't the read range on your floor. Start with one storage location and one type of item, confirm tags stay on, read reliably and that staff are willing to use them, then expand. At the stone plant we also started with a digital-inventory pilot; only after it proved fast and accurate for stocktaking and finding items did we bring RFID into the ERP's location management. The full story is in how a stone plant can take inventory with RFID.

Pre-implementation checklist:

  • Is the goal stocktaking, finding items, or recording movements?
  • How far away must tags be readable? Write down a number.
  • Is the ID format defined? Will tags be reused?
  • How will tags be attached for each item type and storage environment?
  • What happens when a tag falls off — is there a barcode backup?
  • After reading, which system does the data go to?

If your plant has inventory that's hard to count or items that are hard to find — stone industry or not — see our Industry ERP / RFID service, or tell us about your site.

Let's talk