Compliance 📅 September 6, 2026 ⏱ 12 min read

RFID for Returnable Transport Packaging: What the EU's Reuse Targets Mean for Crate and Pallet Pools

Stacked returnable plastic crates and a steel IBC cage passing a four-antenna UHF RFID dock door portal at a pooling depot

A reuse mandate is an asset-tracking mandate wearing different clothes. A pool reaches 40% reusable when it can say, on any given Tuesday, how many crates are at which customer, how many are in the wash, and how many are due back this week. The European Union has now put a date on that requirement, and the date is close enough that hardware decisions taken in 2026 are the ones that will be audited in 2030.

We build UHF RFID readers, gates and tag hardware in New Delhi and ship them to pool operators, poolers’ customers and contract logistics providers across the EU, the Gulf, Africa and Southeast Asia. This article is written from that side of the table: what Regulation (EU) 2025/40 actually says, what a rotation costs when assets leak, where the read points physically go, and which tag suits a caustic wash line on a plastic crate versus a galvanised IBC cage. Where a number appears below, it comes from a source we read and cite.

What Regulation (EU) 2025/40 requires, and when

The Packaging and Packaging Waste Regulation, Regulation (EU) 2025/40, replaces the old packaging directive with a directly applicable regulation — no national transposition, the same text in every Member State. The European Commission states that it “entered into force on 11 February 2025 and applies generally from 12 August 2026”. That second date has already passed. The reuse targets themselves are dated later, which is exactly why the operational work belongs in the current planning cycle rather than the 2029 one.

Article 29 is the article that matters for transport packaging. From 1 January 2030, an economic operator using transport packaging — or sales packaging used to transport products, including e-commerce — within the Union must ensure that at least 40% of that packaging in total is reusable packaging within a re-use system. Article 29 then sets a 70% figure from 1 January 2040 as an endeavour target: an aspiration written into the text rather than the hard obligation that lands in 2030. Plan the fleet against the binding 40%, and read the 70% as the direction of travel the legislator has signalled.

Sitting alongside the 40% figure is a much sharper obligation. For movements between an operator’s own sites, and for deliveries to another operator inside the same Member State, the requirement from 2030 is 100% reusable transport packaging. That is the clause that turns an internal shuttle network, a plant-to-DC lane or a co-packer loop into a closed pool by law rather than by preference.

Article 29 carries carve-outs of its own, and mapping them against your fleet is the first hour of work worth doing. Cardboard boxes sit outside the transport-packaging targets under Article 29(4)(d). So do packaging for dangerous goods and custom large-scale transport packaging. There is also a micro-enterprise exemption, for operators placing no more than 1,000 kg of packaging on the market in a Member State in a calendar year. On 25 February 2026 the Commission added pallet wrapping and straps to that list, adopting a delegated act that removes them from the 100% intra-operator requirement, following a feedback period that ran from 10 December 2025 to 9 January 2026; a Commission study had concluded that exclusive use of reusable wrappings and straps could lead to disproportionate adaptation costs. Plastic crates, trays, foldable plastic boxes, pallets, IBCs, pails, drums and canisters remain inside the 100% rule, and they are the formats an RFID programme is built around.

Date	What changes
11 February 2025	Regulation (EU) 2025/40 enters into force
12 August 2026	The Regulation applies generally across the EU
25 February 2026	Delegated act adds pallet wrapping and straps to the formats outside the 100% intra-operator rule
1 January 2030	At least 40% of transport packaging reusable within a re-use system; 100% for movements between an operator’s own sites and to another operator in the same Member State
1 January 2040	70% endeavour target for transport packaging

One practical instruction before any of this reaches a contract: read the article text on EUR-Lex yourself, in the language version that governs your agreements, before you write a percentage or a date into a supply clause. Definitions such as what constitutes a re-use system, how “in total” is computed across your portfolio, and exactly which of your formats fall under an exemption are all decided by the wording, and your legal team will want the wording.

Which assets are in scope

Article 29 names the formats explicitly: pallets, foldable plastic boxes, boxes, trays, plastic crates, intermediate bulk containers, pails, drums and canisters, of any size or material, including flexible formats, plus pallet wrappings and straps used to stabilise and protect palletised goods — with cardboard boxes carved out under Article 29(4)(d). For an RFID engineer, what remains is not one problem. It is four, because those objects behave in four different ways in a UHF field.

Plastic crates, trays and foldable boxes. RF-transparent when empty, and something else entirely when full. A crate of chilled meat, fresh produce or bottled liquid detunes and absorbs. The crate that sets your real read rate is the wet, loaded one sitting third from the bottom of a stack, so commission the portal against that crate and the figures hold in service.
Pallets. Wood is mostly benign but hygroscopic — a pallet that has stood in the rain is a wetter dielectric than the one you tested with, so test both states. Plastic pallets are consistent. The block or stringer gives you a protected recess, which is the single best gift the geometry offers you.
IBCs and steel drums. A galvanised cage or a steel drum wall is a ground plane, and on-metal construction is what turns that surface into part of the antenna system. Specify on-metal explicitly in the enquiry and the quotation comes back against the right tag family first time.
Pails and canisters. Small, curved, often stacked nested. Curvature and nesting both shape the read zone, so mount under the rim or on the handle boss, where the tag stays clear of the pail above it.

Then there is the boundary question. In law, a pool boundary is where the re-use system says it is. In practice it is wherever an asset is scanned, and the design work is deciding early which movements you will observe and which you will infer from a stated rule. The movement most worth an explicit decision is a customer’s internal one — goods-in to a production line to a satellite store — because it happens entirely inside someone else’s site. Choose whether that leg gets a read point or a modelled status, write the choice into the reporting definition, and the register stays clean, because an asset is then located by either an event or a rule you can show an auditor. That decision shapes everything downstream on a returnable crate and cylinder tracking system.

The economics: why pools leak, and what tracking changes

The compliance argument answers “why must I”. The pool arithmetic answers “why would I want to”, and it is usually the one that releases the budget.

Published figures put the annual loss rate on untracked returnable assets in a wide band. GPX cites a 5% to 15% annual loss rate on returnable assets. Mordor Intelligence, in its market study on returnable transport packaging with IoT tracking, reports that cellular and GPS sensors reduce loss and theft rates by as much as 60% through geofencing every returnable transport item. The same study sizes that tracked-RTP segment at USD 2.72 billion in 2025, growing to USD 4.55 billion by 2030 at a 10.84% CAGR, with pallets holding 34.57% of the market by packaging type in 2024. Those are analyst and vendor figures, published at market level. The number that moves a business case is the one your own dispatch-and-gate data produces in a quarter, and that is the measurement we set up first.

What you can compute exactly is your own model. Here is the arithmetic on a mid-sized crate pool, with the assumptions stated so you can substitute your own.

Input	Value used	Consequence
Fleet size	100,000 crates	—
Replacement cost per crate	€9 (substitute yours)	€900,000 fleet value
Rotations per crate per year	6	600,000 rotations served
Annual loss at 5%	5,000 crates	€45,000 replacement spend
Annual loss at 15%	15,000 crates	€135,000 replacement spend
Value of one percentage point of loss	1,000 crates	€9,000 per year, recurring

The spread between the two ends of that published loss band is €90,000 a year on a €900,000 fleet — ten percent of the asset base, annually, decided by how well you know where things are. Against that, a rugged crate tag is a single-digit USD item and a four-antenna dock portal is a four-figure USD one. We quote in USD on EXW, FOB, CIF or DDP terms, and we price the read points the rotation actually needs, which is usually fewer than a floor plan suggests.

Then there is the saving nobody budgets for. Assets that are merely unfound still consume capital, because a planner covering peak demand treats them the same as assets that are gone. If 4% of a 100,000-crate fleet is unlocated at any moment, the safety stock bought to cover that gap is 4,000 crates — €36,000 of purchasing against inventory that already exists and is sitting on a customer’s mezzanine. Locating that proportion releases working capital immediately, before a single genuinely lost crate is recovered. In practice that second-order saving lands sooner than the loss-rate improvement, because it needs only visibility, not behaviour change.

Where the read points go

A rotation is defined by events, and each event needs a place where it can be observed. Five candidate read points cover the great majority of pool movement.

Dispatch door. What left, and against whose order. This is the event that creates the liability, and it is where a deposit or hire charge legitimately starts.
Customer receipt. Confirms delivery and closes the dispute window. Where the customer will host a reader, this is the highest-value point in the chain. Where a customer prefers not to host fixed hardware, an app-driven handheld read at the tailgate captures the same event, timestamped and attributed to a named operator, using the Android, Windows or Flutter SDK.
Customer return / collection. The counterpart event. Receipt and return together define the completed rotation, which is the unit everything else is measured in.
Wash line and repair loop. Condition and cycle count. A read on entry to the washer and a read on exit gives you cycles-since-new per individual asset, which is what lets you retire crates on evidence instead of on a fleet-wide age assumption.
Depot gate. The reconciliation point. Everything the doors missed shows up here, because the gate sees the vehicle rather than the pick.

The minimum viable set is dispatch door plus depot gate. That pair alone tells you what your estate believes it issued and what physically crossed the boundary, and the difference between those two numbers is the most useful diagnostic in the whole system. Add the wash line second, because it is a controlled, single-file, low-speed environment that delivers near-perfect reads for very little engineering — it is the cheapest high-quality data in the building. Customer-side read points come third, when a lane has enough volume to justify installing at someone else’s address. The staging logic and the put-away side of this normally lands in an existing warehouse management system rather than a new screen.

Tag selection by substrate

Tag choice on returnable assets is a mechanical engineering problem at least as much as an RF one. The tag has to serve the asset’s full service life, which for a plastic crate means hundreds of wash cycles, forklift impacts, stacking loads and UV.

Asset	Mounting	Tag construction	What sets read range
Plastic crate / tray	Moulded pocket or recess in the end wall; rivet or weld-on plate as a retrofit	Rigid encapsulated hard tag, IP-rated, chemical-resistant housing	Wet or dense product detuning the antenna; depth of the recess; position in the stack
Foldable plastic box	Recess on the fixed base, clear of the hinge line	Slim rigid tag, low profile so folded stacks still nest	Folded walls sitting directly over the tag
Wooden or plastic pallet	Block or stringer recess, two tags on opposite sides for orientation tolerance	Rugged pallet tag, impact-resistant	Moisture in wood; the load itself, once the pallet is full
Steel IBC cage	Bolted or riveted to the cage frame, or on the composite base pallet	On-metal tag with integral spacer	Standoff distance; the cage geometry in each approach orientation
Steel drum	Adhesive or weld-on plate on the chime or side wall	On-metal tag, curved-surface tolerant	Curvature and the ground-plane behaviour of the drum wall
Pail / canister	Under the rim or on the handle boss	Small-footprint on-metal or hard tag	Nesting; small antenna aperture; orientation

Two engineering points decide whether a returnable-asset deployment holds its read rate for the life of the fleet.

Design the zone around the range the asset actually delivers. An on-metal tag works by putting a controlled spacer between the antenna and the conductive surface, and that spacer is what makes a steel drum or an IBC cage readable at all. The working range it achieves on your drum, in the orientation it presents at the door, is the figure to build the read zone around — measured on your own asset, wet and loaded, because that is the figure the commissioning test will reproduce. Our datasheets state the free-space reference so you can compare tag families on a common basis, and we will run the loaded measurement with you on sample assets before you commit to a fleet quantity. Our guide to pallet-level tracking works through the same trade-off on unit loads.

State the environment as a specification. Before anyone quotes, write down the wash temperature, the cycle time, the detergent chemistry and pH, the peak steam exposure, the sanitiser used, and the IP rating you need. Then require the supplier to answer against those figures. Tags are rated against a stated temperature and chemistry window, so a tag specified against your actual wash line holds its read rate for the life of the asset. Writing the wash parameters into the enquiry is what secures that, and it is a two-line addition to a specification.

Read-zone design at a depot door and on a conveyor

Stacked loads are the honest test of a portal. The tag on the outside of the stack always reads. The one in the geometric centre, surrounded by wet product on five sides, is the one that decides your actual read rate — and therefore your reconciliation accuracy. Specify and commission against that tag, with a full pallet of your real product, not with a demonstration stack.

Antenna geometry. Work the footprint from the beamwidth rather than guessing. As an illustrative worked model, with every input stated so you can substitute the antenna you are actually quoting: take a circularly polarised panel with a 70° half-power beamwidth, mounted 2.6 m above the floor and tilted 20° down. Its boresight meets the floor about 7.1 m out (2.6 / tan 20°), while the lower −3 dB edge, at 55° below horizontal, lands about 1.8 m out (2.6 / tan 55°). The illuminated strip is therefore roughly 1.8 m to well past 7 m — wide, and deliberately so — and the upper half of the beam never meets the floor at all. It carries straight down the aisle, and that is precisely the energy to manage if you want clean separation between zones. A four-antenna door — two per side, one low for base-of-stack tags and one high, angled inwards — is the configuration that survives commissioning most reliably. Re-run the same two lines of trigonometry with your own mounting height, tilt and beamwidth before the drawings are signed off.

Dwell time. A forklift at 6 km/h moves at 1.67 m/s. Through a 2 m deep read zone that is 1.2 seconds of presence. With four antenna ports cycling, each port gets roughly 300 ms of transmit time per pass, and that is your entire budget for singulating every tag on the load. Raise the approach to 8 km/h and the budget falls to about 900 ms total. Either put a speed limit on the approach or add zone depth; those are the two levers, and both are cheap when they are decided at design stage.

Session flags. The RAIN Alliance system design guidelines set out the logic clearly. Session S0 suits small populations of fast-moving tags — conveyors and tolling — because tags reset to A when de-energised. Their worked example: around 20 tags in the zone at 100 tags per second throughput needs 200 ms, inside the FCC 400 ms channel dwell limit. Their counter-example takes 50 tags in the same conditions, where S0 stops being the right choice under FCC rules because the tags reset on every frequency hop. Under ETSI EN 302 208 the channel dwell allowance is up to 4,000 ms, which gives S0 far more room in Europe. For a stacked pallet at a door, where the population is large and slow-moving, S1 (minimum 500 ms persistence) or S2 is the right choice. A wash-line conveyor, single file and fast, is the textbook S0 case.

Spill-over. Test for it before it reaches a reconciliation, because a missed read announces itself as a gap while a cross-read arrives looking like a fact. The test is simple. Park a full, tagged load at door 1. Run door 2 with nothing moving through it. Count the reads. Anything above zero is worth attention — more down-tilt, lower power, absorber panels, or reader-side zone gating on RSSI and read count. In a dense installation, separate the transmit channels: the RAIN guidelines recommend operating readers on channels 4 and 10, and channels 7 and 13, when readers operate in a dense, close environment in the European lower band.

On band selection for export, our UHF readers are configured per order for the ETSI band (865–868 MHz) or the FCC band (902–928 MHz). India’s licence-free allocation now mirrors the ETSI lower band under G.S.R. 853(E) of 10 December 2021 — 2 W e.r.p., channels of 200 kHz or less, permitted on the four channels centred at 865.7, 866.3, 866.9 and 867.5 MHz, with continuous transmission capped at 4 seconds and at least 100 ms between transmissions on a channel, referenced to EN 302 208. Equipment type-approved under the superseded 2005 rules stays valid for its life.

Proving reuse: the evidence an auditor will ask for

A reuse claim is only as good as the record behind it. Build the record deliberately, because it is far easier to design the evidence in than to reconstruct it from event logs written for something else.

An asset register. One row per physical asset, keyed on the tag’s unique identifier, carrying date of entry into service, format, owner and current status. This is the spine: a register is what turns a stream of reads into an inventory.
A written definition of a completed rotation. Auditors will ask what you counted. Ours is normally dispatch to customer receipt to return to depot, with the wash read as the event that returns the asset to available stock. Whatever definition you choose, write it down once and apply it consistently, because a stable definition is what makes one year’s figures comparable with the next.
A closed reconciliation. Every asset sits in exactly one bucket — issued, at customer, returned, in wash, in repair, retired, or unaccounted — and the buckets sum to the register. Report the unaccounted bucket openly rather than absorbing it. A pool that publishes its unaccounted figure internally is a pool that can drive it down, quarter by quarter.
Retention and export. Keep event-level data for a defined period and be able to export it as flat files with timestamps, asset IDs, read points and originating reader identity. A claim that a third party can export and recompute is evidence; the export capability is what converts an internal number into an auditable one.

The architectural point that decides whether the record is complete: hold events at the edge as well as centrally. Our readers buffer reads locally and forward them when the link returns, so a depot that loses its uplink for three hours produces a delayed record rather than a missing one. Keep the durable, append-only event stream centrally, and keep enough local storage on each reader to cover the longest outage you have actually experienced, plus a margin. A reuse record is built forward, which is why the event stream goes in before the first audit period opens.

Integrating with the pool operator's systems

The reader network produces events. Value appears when those events land in the systems that already run the business.

ERP and TMS. Map each read point to a business event: dispatch, receipt, return, hire, off-hire. Where a shipment already exists as an order, the RFID read should confirm or correct it rather than create a parallel record — a single source of truth per movement is the integration decision that pays back most reliably. Events flowing into an existing logistics and supply chain platform generally beat a separate pool portal that nobody logs into.

Billing on scans. Deposit and rental models get materially better when the clock starts and stops on a physical read rather than on a delivery note keyed in two days later. It shortens the dispute cycle, because the answer to “we returned those” becomes a timestamp and a gate identity. This is often the change that pays for the installation, ahead of any loss-rate improvement.

Customer-facing visibility. Give each customer their own holdings, ageing and return history, and nothing else. A customer who can see 400 crates ageing past 30 days on their own account will usually return them without being chased, which is a cheaper collection mechanism than an invoice.

Phasing. One depot, one lane, one asset class. Instrument the dispatch door and the depot gate, run for a quarter, and compare what the system says left with what the customer says arrived. That comparison teaches you more about your pool than any pilot report, and it gives you a measured baseline loss figure — your own, not a published range — to justify the rollout. Then add the wash line, then the second asset class, then the customer sites with the highest volume.

Our UHF reader models carry WPC ETA and BIS registration for India, which matters when the pool operator is EU-headquartered but the depots are in Chennai or Pune and one hardware standard has to serve both. Hardware and software are built in-house in New Delhi, so a reader firmware change and the middleware that consumes the events come from the same engineering team.

The 2030 dates are fixed and the intra-operator obligation is absolute for the formats in scope. The pools that will find that comfortable are the ones already counting.

Frequently asked questions

What does the PPWR require for reusable transport packaging?

Article 29 of Regulation (EU) 2025/40 requires that, from 1 January 2030, at least 40% of transport packaging used within the EU is reusable packaging within a re-use system. A 70% figure follows from 1 January 2040 as an endeavour target rather than as the hard obligation that lands in 2030. For movements between an operator’s own sites, and deliveries to another operator within the same Member State, the requirement from 2030 is 100% reusable transport packaging. Article 29 carries carve-outs of its own — cardboard boxes under Article 29(4)(d), dangerous-goods packaging, custom large-scale transport packaging, and micro-enterprises placing no more than 1,000 kg of packaging on the market in a Member State in a calendar year — and a delegated act adopted on 25 February 2026 added pallet wrapping and straps to that list for the 100% requirement.

When does Regulation (EU) 2025/40 start to apply?

The European Commission states that the Regulation entered into force on 11 February 2025 and applies generally from 12 August 2026. The reuse provisions in Article 29 carry their own later dates — 1 January 2030 for the binding 40% and 100% figures, and 1 January 2040 for the 70% endeavour target — while the underlying obligations, definitions and system requirements are already live. That is why fleet tagging and read-point design belong in current planning rather than in 2029.

Which items count as reusable transport packaging?

Article 29 lists pallets, foldable plastic boxes, boxes, trays, plastic crates, intermediate bulk containers, pails, drums and canisters, of any size or material, including flexible formats, together with pallet wrappings and straps used to stabilise palletised goods. Cardboard boxes sit outside the Article 29 targets under Article 29(4)(d), as do dangerous-goods packaging and custom large-scale transport packaging, and the February 2026 delegated act added pallet wrapping and straps to the formats outside the 100% intra-operator requirement. Plastic crates, trays, foldable plastic boxes, pallets, IBCs, pails, drums and canisters remain in scope. Confirm scope against the article text on EUR-Lex before writing any figure into a supply contract.

Does the PPWR require RFID?

The Regulation sets reuse outcomes and requires that reusable packaging operates within a re-use system; it specifies the result rather than the identification technology. RFID becomes relevant because meeting a 40% reuse share and evidencing it needs an asset register, rotation counts and a closed reconciliation across issued, returned, in-wash, in-repair and unaccounted stock. UHF RFID is the practical way to generate those events at dock doors, wash lines and depot gates while keeping every movement hands-free.

What RFID tag works on a plastic crate that goes through a wash line?

A rigid, encapsulated hard tag with a stated IP rating and chemical resistance, fitted into a moulded pocket or recess in the crate wall, or riveted to a plate as a retrofit. Specify the wash temperature, cycle time, detergent chemistry and pH, steam exposure and sanitiser as requirements before you request a quotation, and have the supplier answer against those figures. Validate read range with the crate wet and loaded with your real product: that measured figure is the one to design the read zone around, and it is the figure the commissioning test will reproduce.

How do you track IBCs and steel drums with RFID?

Both call for on-metal tag construction, which uses a controlled spacer between the antenna and the conductive surface so the metal becomes part of the antenna system. On an IBC, the usual mounting points are the cage frame or the composite base pallet; on a drum, the chime or side wall, with a tag tolerant of the curvature. On-metal construction is what makes a steel drum or an IBC cage readable at all, and the working range it delivers on your own asset, in its loaded state and in the orientation it presents at the door, is the figure to design the read zone around.

How do you measure crate loss rate with RFID?

Instrument the dispatch door and the depot gate first, then reconcile: assets issued, minus assets returned, over a defined period, with in-wash and in-repair stock excluded from the unaccounted bucket. Every asset in the register sits in exactly one status and the buckets sum to the register total. Published figures put annual loss on untracked returnable assets at roughly 5% to 15%, and the number that justifies investment is your own measured baseline, which one quarter of dispatch-and-gate data will give you.