Pallet-level tracking tells you a consignment moved. Item-level tells you which units are inside it, and that is a different engineering problem — thousands of tags per facility, custody changing hands several times, and no single party owning the infrastructure end to end. It is also where DECT NR+ becomes interesting, because the density and the absence of an operator are exactly the constraints it was standardised for.
Tracking a container is straightforward. One tag, one asset, a satellite or cellular link, and a position every few hours. Tracking the four thousand individual units inside that container is a different discipline entirely, and most of the difficulty has nothing to do with positioning.
Density is the first constraint. A distribution centre holding high-value electronics or pharmaceuticals may have tens of thousands of tagged units within a few hundred metres. Any protocol whose collision behaviour degrades with population, or which needs a gateway per few hundred devices, becomes an infrastructure problem before it becomes a data problem.
Custody is the second. A unit passes from manufacturer to freight forwarder to port to carrier to distribution centre to final delivery. Each hop is a different company, with different systems, different infrastructure and no obligation to install anything for the benefit of the previous party. The tag has to keep working across all of them and produce a record all of them can agree on.
And the value changes the threat model. At item level for high-value goods, the adversary is not weather or coverage — it is diversion, substitution, grey-market resale and theft in transit, often by someone with legitimate access to the consignment.
No option wins outright. The choice follows from density, who owns the infrastructure, and whether the tag must survive a handover.
| Technology | Where it fits | Where it breaks down |
|---|---|---|
| Passive RFID (UHF) | Very low tag cost, no battery, portal and bulk reads at choke points | Read at points, not continuously; metal and liquid degrade reads badly; no sensing and no tamper detection |
| BLE tags | Low cost, multi-year coin cell, dense deployments, sensing possible | Gateway infrastructure per zone; advertising collisions at very high density; no native long-range relay |
| UWB | Sub-metre position where the operational question is which bay, not which room | Anchor network required and surveyed; higher energy per fix; cost per point rules out true item level |
| DECT NR+ (DECT-2020 NR) | Autonomous mesh with no operator and no base station, in licence-exempt spectrum, designed for very high device density | Regional spectrum availability varies; a younger silicon ecosystem; relaying nodes carry a heavier energy burden |
| Cellular (LTE-M / NB-IoT) | Wide area with no site infrastructure at all | A subscription per device makes item level uneconomic; power cost per message is high for dense reporting |
At item level the deciding factors are usually cost per tag at volume and whether the receiving party has to install anything. Both favour approaches that do not require an operator relationship.
DECT-2020 NR, also called NR+, is an ETSI standard that the ITU recognised as meeting IMT-2020 requirements — making it the first non-cellular technology admitted as 5G. For item-level logistics its relevant properties are structural rather than headline.
It has no base station. Nodes form a decentralised mesh, discover each other and relay autonomously. A tag deep inside a loaded trailer, with no path to any fixed infrastructure, can reach the outside world through the tags around it. That single property addresses the case that defeats star-topology networks.
It has no operator. It runs in licence-exempt DECT spectrum, so there is no subscription, no SIM and no commercial relationship required to switch it on inside your own facility or vehicle. At tens of thousands of units, a per-device subscription is usually the argument that ends an item-level project before it starts.
It was specified for density. The standard targets device counts per square kilometre far beyond what short-range star networks handle gracefully, which is the regime an item-level deployment actually operates in.
The honest counterweights: spectrum allocation in the DECT band is not uniform worldwide and has to be confirmed per market; the silicon ecosystem is younger and narrower than BLE or LoRa; and mesh relaying is not free — nodes that carry other nodes' traffic drain faster, so routing policy becomes part of the battery budget rather than an afterthought.
Tags associated with unit identity at manufacture or pack-out, not on arrival. An association performed later is a record of what someone believed was in the box.
The useful record is who held it and when it changed hands, with position as supporting evidence. Chain of custody is what a dispute is settled on.
Each party has different readers, different systems and no obligation to the last. The tag must produce a record that survives the handover, and the platform must reconcile evidence from parties who do not trust each other.
For high-value goods, an intact seal matters as much as a location. Opening detection, light sensing inside sealed packaging and a record that cannot be quietly rewritten after the fact.
Temperature and shock for pharmaceuticals and precision electronics, recorded continuously and tied to the same unit identity as the custody record.
Ten thousand units reporting normally is not information. A unit that left the geofence, broke a seal, exceeded a temperature or failed to arrive is — and it has to reach someone with authority to stop the shipment.
Establish the unit value threshold and who owns infrastructure at each hop before selecting a link layer. Those two answers eliminate most of the option space, and they are commercial questions rather than engineering ones.
Tracking individual units rather than the pallet, case or container holding them. It differs from asset tracking mainly in density — tens of thousands of tags in a facility — and in the fact that custody changes hands several times between parties with different infrastructure.
DECT-2020 NR, or NR+, is an ETSI standard recognised by the ITU as meeting IMT-2020 requirements, making it the first non-cellular 5G technology. It forms an autonomous decentralised mesh with no base station and no operator, in licence-exempt spectrum, and was specified for very high device density — which matches item-level deployment closely.
No. It runs in licence-exempt DECT spectrum with no subscription and no SIM. At tens of thousands of units, a per-device subscription is usually what ends an item-level business case, so removing it changes the arithmetic materially.
Spectrum allocation in the DECT band is not uniform worldwide and must be confirmed per market; the silicon ecosystem is younger and narrower than BLE or LoRa; and mesh relaying costs energy, so nodes carrying other nodes' traffic drain faster and routing policy becomes part of the battery budget.
It answers whether a unit passed a point, at very low tag cost and with no battery. It does not provide continuous custody, condition monitoring or tamper evidence, and metal and liquid degrade read reliability considerably — so for high-value goods it is usually one layer rather than the whole system.
Rarely the radio. Tag cost against unit value, nobody owning infrastructure in the middle of the journey, tags that can be moved between units undetected, alert volume at scale, and the gap between operational visibility and a compliant regulatory record.
Written for engineers. Share it with one.
Indoor, outdoor and out-of-coverage location.
KNOWLEDGERange, power and infrastructure compared.
PRODUCTPortal and zone reads at low tag cost.
PRODUCTZone-level presence with multi-year battery.
PRODUCTSub-metre where the bay matters.
INDUSTRYInventory accuracy and shrinkage at estate scale.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.