How to Evaluate RFID Tracking for Manufacturing Supply Chain Visibility

Our last post argued that an RFID read improves supply chain visibility only once business context is attached to it. For a team weighing an RFID program, that argument becomes three questions: which movements you need to see, what each read will tell you, and who will act on it.

Pick the movements that matter

Start with one asset group or one movement path, and name what it costs you not to see it. For returnable containers moving between plants and partners, the cost shows up as searches, replacement purchases, and production lines waiting on equipment. Between 15 and 30% of a returnable fleet goes missing or gets damaged each year, so replacement spend is the first number to pull. For materials crossing warehouse handoffs, the cost shows up as inventory discrepancies found after they have already disrupted a shipment.

Write down which missing event causes each cost. A replacement purchase can trace back to a return nobody recorded. A discrepancy can trace back to a receipt logged late or not at all. Then define success in those terms before anyone chooses a tag: fewer replacement purchases, fewer hours spent searching, dwell time at partner sites that you can measure.

Decide where RFID fits the workflow

RFID earns its place at repeated, high-volume points where manual scanning is hard to sustain: dock doors, conveyors, the gate where returnables come back. A fixed reader there records every tagged item that passes without anyone stopping to scan it. Place readers where a read answers one of the questions you wrote down, at departure, arrival, transfer or return. For a multi-site example, see how RFID tracks returnable assets across sites.

Two limits belong in the evaluation. A reader records an item only when it passes a read point, so RFID tells you where something was last seen and when, with no location data between points. UHF RFID also reads less reliably around metal and liquids, so steel racks and filled containers need tags built for them and testing in the real environment. At some steps, a barcode scan or a mobile device will be the more reliable and cheaper choice.

Evaluate what sits behind each read

A read records a tag identifier, a reader location, and a time. Ask a vendor to show you, with your own data, what the system does next.

Can it tie the tag to the right asset, product, or material, including after a damaged tag is replaced? Can it attach the event to an order, a lot, a partner, and a business step such as shipping or receiving? GS1’s EPCIS standard describes supply chain events this way: by what, when, where, and why, so ask whether a system records events in EPCIS form, which lets other systems and partners read them. Can it compare the movement with the plan and flag the difference, such as a container on the wrong route, an overdue return, or a dwell time past its limit?

Then follow the exception. Decide which system receives it (the ERP, the WMS or a partner portal) and how fast it has to arrive to be useful. A missed return flagged a week later goes into a report. Flagged the same day, it becomes a call to the partner holding the container. Name the person who investigates a missing return or an unexpected movement before go-live.

Run a focused pilot

Pilot one asset group on one path. Before the first tag goes on, measure the baseline for the costs you listed at the start: replacement purchases per month, hours spent searching, average dwell time at partner sites, and missed or late returns.

During the pilot, test read rates on the actual containers, through the actual dock doors, in the actual operating conditions. Check that exceptions reach the person named to act on them, and that they do act. Judge the pilot by the change in the baseline numbers.

A US bakery that tracks its returnable trays with ACSIS shows why the unit cost matters. Its trays and dollies were untagged, so there was no way of tracking where or why they went missing. At about $7 a tray, every missing tray had a cost the company could count. After tagging trays and dollies with RFID and barcodes linked to the ACSIS Cloud Platform, with KPIs and automatic alerts on asset performance, the bakery reported a 33% reduction in costs, with per-tray savings in the first year.

ACSIS, part of Antares Vision Group, captures physical events across products, materials, assets, and partners, connects them to business context, and delivers them to the ERP, WMS, and partner systems. Explore how ACSIS tracks returnable assets across facilities and partners.

Have your baseline numbers? Contact ACSIS to scope a pilot on one asset group or movement path.