Barcode is the cheapest and most accurate at item level but needs line of sight and a person scanning. Passive UHF RFID reads many tags at a distance without line of sight and gives zone-level location. BLE beacons give room-level position with cheap battery tags. UWB gives 10–30 cm real-time position at the highest infrastructure cost. Most sites end up using two of them together.
Swedish Technology selects tracking technology from measured site conditions and a written accuracy target, not from a vendor datasheet, and documents the trade-off so the decision survives an audit.
What problem does this solve?
Tracking projects usually start with a symptom, not a technology: the annual asset count takes three weeks, spare parts cannot be found before a shutdown, tools leave the site and never come back, or a customer asks for proof that a pallet was loaded. By the time the requirement reaches procurement it has already been translated into a product name — someone has been shown an RFID gate, a BLE beacon kit or a UWB demo with a moving dot on a floor plan — and the evaluation becomes a price comparison between things that do not do the same job.
The four technologies answer different questions. A barcode answers what is this item when a person points a scanner at it. Passive UHF RFID answers which of these items just passed this point, or are in this zone without anyone scanning. BLE answers roughly where is this asset right now, at room level, for years on a coin cell. UWB answers exactly where is it, to a few centimetres, several times per second. Buying UWB to solve an inventory-count problem, or barcodes to solve a real-time-location problem, is the most common and most expensive mistake we are asked to fix.
The second failure mode is ignoring the physics and the environment. Metal racking, liquids, dense stock, forklift traffic, reinforced concrete and reflective steel walls change read rates and position accuracy far more than the brand of reader. A technology that performs perfectly in a supplier demo room can lose a third of its reads on your dock door, and no amount of software will recover a tag that was never read.
How the solution works
Choose from the question you need answered, then from the environment, then from cost. Write down the operational question in one sentence ("we need to know, within one hour, which of 4,000 tools is inside the workshop"), the accuracy that makes the answer useful, how often it must refresh, and who acts on it. That single sentence eliminates two of the four technologies immediately. Only then do tag prices, reader counts and licence models matter.
In practice most UAE and GCC deployments we build are hybrids: barcode or QR remains the fallback and the item-identity standard, passive UHF RFID handles bulk counting and gate events, and BLE or UWB is added only in the areas where real-time position genuinely changes a decision. Swedish Technology develops the middleware layer that normalises all of them into one event stream and one asset record, so the choice of radio stays an engineering detail rather than a lock-in.
- 1Input — define the question State the operational question, the decision it drives, the required accuracy (item, zone, room, centimetre), the refresh rate and the acceptable miss rate. Anything without a decision attached is reporting, not tracking.
- 2Capture — survey the environment Walk the site: materials the tags will sit on (metal, liquid, cardboard), pallet density, ceiling height, forklift speeds, existing Wi-Fi and cabling, and the RF noise floor. Photograph every candidate read point.
- 3Processing — shortlist by physics Match the environment to the technology: line-of-sight availability, on-metal tag requirements, multipath risk for UWB, anchor density for BLE, and regional frequency rules for UHF RFID.
- 4Trial — measure, do not assume Run a short bench and on-site trial with three to five tag models per asset type at the real read points, recording read rate, false reads and position error against a known ground truth.
- 5Integration — check the receiving system Confirm what SAP, Oracle, Odoo, IBM Maximo or the WMS can actually accept: goods movements, asset status, location codes or only a report. The receiving system often caps the useful accuracy.
- 6Action — cost the full lifecycle Tags per year, readers or anchors, cabling and power, mounting, licences, spares, battery replacement labour and the integration effort — over five years, not at purchase.
- 7Reporting — decide and document Produce a one-page decision record: chosen technology, measured evidence, rejected options with reasons, and the accuracy the business may rely on. This is what protects the project at audit.
Reference architecture
Whichever radio you pick, the layers above it stay the same. Designing them once is what makes a later change of technology survivable.
| Layer | What it contains |
|---|---|
| Identification layer | The tag or label and its identifier scheme — GS1/EPC for RFID and barcode, MAC or vendor ID for BLE, device ID for UWB. One asset should carry one meaningful identity, even if it is encoded on two carriers. |
| Capture layer | Handheld scanners, fixed UHF readers and antennas, BLE gateways or UWB anchors. This is the layer with the physics, the cabling and the site work — and the one that determines your real data quality. |
| Edge / middleware | Filtering, de-duplication, smoothing, direction logic at gates, and translation into business events (received, moved, issued, left site). Without it the ERP receives thousands of meaningless reads. |
| Application layer | Asset register, inventory counts, maps and dashboards — for example MOWQIE for RTLS tracking or a WMS for stock — with rules for exceptions and alerts. |
| Enterprise systems | SAP, Oracle, Odoo, IBM Maximo or a custom asset register, receiving clean business events rather than raw reads, with idempotency so a repeated read never doubles a movement. |
Deployment options: Capture hardware is always on-premise. Middleware and applications run on-premise, in a UAE-region private cloud or air-gapped, depending on the site's classification; only aggregated reporting typically leaves the operational network.
Key capabilities
Technology selection with measured evidence
You choose between barcode, RFID, BLE and UWB from read rates and position errors measured on your own site, not from a datasheet.
availableTag and label trials per asset type
Each asset family gets a tag that survives its environment — on-metal, high-temperature, laundry, outdoor UV or embedded.
availableHybrid barcode + RFID identity scheme
One identifier printed and encoded, so handheld fallback works when a reader is down.
availableGate, portal and zone design
Read points that report direction and avoid stray reads from neighbouring aisles.
custom developmentUnified event middleware
RFID, BLE and UWB events arrive in one normalised stream, so the ERP integration is written once.
custom developmentFive-year cost model
A comparison including tags, batteries, spares, licences and labour, so the cheapest purchase is not mistaken for the cheapest system.
availableMap and digital-twin overlay
Positions and zone occupancy shown on a floor plan or 3D model for supervisors and control rooms.
custom developmentIntegrations
The receiving system usually decides how much accuracy is worth paying for. These are the integration points we implement as custom development.
| System | Integration point & data exchanged | Direction |
|---|---|---|
| SAP MM / PM / EWM | Goods movements, physical inventory documents and equipment status updated from filtered gate and zone events instead of manual entry. → RFID Integration with SAP, Oracle, Odoo & IBM Maximo | bi-directional |
| IBM Maximo | Asset last-seen location and custody, plus work-order triggers when an asset leaves an authorised area. → RFID Integration with SAP, Oracle, Odoo & IBM Maximo | bi-directional |
| Odoo Inventory | Stock moves, lot/serial tracking and cycle-count adjustments from RFID reads via the external API. | bi-directional |
| Warehouse management (Octopus WMS) | Receiving, put-away confirmation and dispatch verification driven by gate reads rather than handheld scans. → Octopus WMS | bi-directional |
| RTLS platform (MOWQIE) | BLE and UWB position streams, zones and geofence alerts for people, vehicles and mobile equipment. → MOWQIE ÔÇô RTLS Tracking | inbound |
| GIS / digital twin (RASM) | Indoor positions and zone occupancy rendered on floor plans or a 3D twin alongside other operational layers. → RASM ÔÇô Digital Twin | inbound |
Industry use cases
Warehousing & distribution
Passive UHF RFID at dock doors for automated receiving and dispatch verification, barcode retained for exception handling and item picking.
Government asset registers
Passive RFID for annual IT and furniture inventory across floors and buildings, cutting a multi-week count to a walk-through with a handheld.
Oil, gas & heavy industry
On-metal RFID for equipment and tool cribs; UWB in a turnaround area where contractor position and exclusion zones must be enforced in real time.
Healthcare & laboratories
BLE tags on mobile equipment for room-level search; RFID for consumable stock; barcode for patient and sample identity where regulation requires it.
Construction & yards
BLE or RFID on plant and formwork for yard-level location and site-exit alerts, where centimetre accuracy would add cost without changing any decision.
UAE & GCC considerations
UHF RFID equipment used in the UAE must operate in the frequency band and power limits set by the national regulator (TDRA), which follows the European allocation rather than the North American one — imported readers configured for another region will underperform or be non-compliant. Government and defence sites frequently require the middleware and application tiers to run on-premise or in a UAE-region private cloud, with Arabic and English interfaces for operators and bilingual hand-over documentation. Procurement typically expects a documented site survey and a proof of concept before a framework purchase, so evidence from a measured trial is worth more than a technology brand during evaluation.
Implementation approach
- 1Requirement statement (days) One sentence per operational question, with the decision, accuracy, refresh rate and owner. Signed by the operations manager who will use the data.
- 2Asset and environment inventory Asset families, quantities, materials, movement patterns and every candidate read point, with photographs and a marked-up site plan.
- 3Shortlist and cost model Two candidate designs with five-year costs, infrastructure counts and the risks of each, including the do-nothing option.
- 4Site survey and bench trials RF noise measurement, tag trials per asset family, reader and antenna placement tests at the real read points.
- 5Proof of concept One zone or one gate instrumented for two to six weeks against written success criteria and a ground-truth count.
- 6Decision record Chosen technology, measured results, rejected options and reasons, integration scope and the accuracy the business may rely on.
- 7Rollout design Tag application process, labelling standard, reader and anchor layout, middleware rules, ERP interface specification and operator training.
- 8Hand-over As-built documentation, spare parts, admin guide, source code for custom components, and a maintenance plan including battery replacement schedules.
Security & deployment
Tracking data is operational data about people, vehicles and valuable assets, so it is treated as sensitive by default. Readers and anchors sit on a segmented operational network with no direct internet route; the middleware holds the only credentials to enterprise systems and writes an audit log of every event it forwards. Passive RFID tags broadcast their identifier to anyone with a reader, so identifiers should be meaningless outside your database and, where the assets are sensitive, tag data should be minimal or use the memory-locking and kill features of the EPC Gen2 standard. Where the site is classified, the full stack — capture, middleware, database and dashboards — runs on-premise or air-gapped, with export limited to reports.
Limitations & prerequisites
- No radio technology reaches 100 % read rate in a real environment; a design target is typically 98–99.5 % at a gate, and the exceptions process matters as much as the technology.
- Passive UHF RFID performance drops sharply on metal and liquid-filled items unless on-metal or spacer tags are specified and tested per asset family.
- BLE gives room or area accuracy, not centimetres; claims of sub-metre BLE accuracy usually assume dense gateways, a calibrated environment and a static tag.
- UWB accuracy degrades with obstructions and multipath, needs line of sight between tag and several anchors, and requires cabled or powered anchors — cost grows with area, not with tag count.
- Battery-powered tags have a service life; a 5,000-tag BLE estate is a recurring replacement programme, not a one-off purchase.
- Barcode remains the most accurate item-level identification, but every read costs a human action, so it does not scale to continuous visibility.
- The accuracy of any system is capped by the process: if tags are applied inconsistently or removed at packing, no technology recovers the missing data.
Barcode vs passive UHF RFID vs BLE vs UWB
Indicative figures for planning. Real range and accuracy depend on tags, antennas, materials and the site, and should be measured during a survey.
| Criterion | Barcode / QR | Passive UHF RFID | BLE | UWB |
|---|---|---|---|---|
| Typical read range | 5–50 cm, aimed | 1–10 m (gate/handheld) | 10–50 m to a gateway | 10–50 m to an anchor |
| Location granularity | The scan point only | Zone, gate or shelf | Room / area, 1–5 m | 10–30 cm, real time |
| Line of sight | Required | Not required | Not required | Preferred; accuracy drops without it |
| Tag cost | Printing cost only | Low per tag; higher for on-metal | Medium | High |
| Battery in the tag | None | None | 1–5 years, replaceable | 1–3 years, duty-cycle dependent |
| Infrastructure | Scanners and printers | Readers, antennas, cabling at read points | Gateways every 15–30 m | Anchors with power/PoE and timing sync |
| Calibration effort | None | Antenna tuning and power per gate | Signal mapping per area | Anchor survey, sync and site calibration |
| Reads many items at once | No | Yes, hundreds per second | Yes (beaconing) | Yes, limited by update rate |
| Best fit | Item identity, verification, low volume | Bulk counting, gates, inventory | Search and room-level visibility | Safety zones, process accuracy, high-value tracking |
A common, defensible design is barcode for identity, passive RFID for counting and gates, and a small UWB or BLE zone only where real-time position changes an operational decision.
FAQ
Rarely in full. RFID removes the need to aim a scanner and reads many items at once, but barcodes remain the cheapest fallback, are readable by any partner, and are often mandated on shipping labels. Most projects encode the same identifier in both, so a handheld scan still works when a reader fails or a tag is damaged.
Barcode cost is labour per scan. RFID cost is tags multiplied by item count plus readers per read point. BLE cost is tags plus gateway density plus battery replacement. UWB cost is driven by area: anchors, power and cabling scale with square metres, and the tags themselves are the most expensive of the four.
With a normal gateway density expect room or area level, roughly 3–10 m. With dense gateways, fingerprinting or Bluetooth direction finding it can reach 1–3 m for slow-moving tags. If your process needs to know which of two adjacent bays an item is in, plan for UWB or a physical read point instead.
Yes, and it is usually the right answer. The middleware normalises reads from handhelds, fixed RFID readers, BLE gateways and UWB anchors into one event model, so the ERP or asset register sees consistent events regardless of the radio that produced them.
A requirement statement and shortlist take one to two weeks. A site survey with tag trials adds one to two weeks. A proof of concept on one zone or gate runs two to six weeks. Most organisations can make a defensible decision within six to eight weeks.
Yes. Readers, anchors, middleware, database and dashboards all run inside your network. Only optional analytics or remote support need outbound connectivity, and both can be omitted for classified sites.
An asset or SKU list with quantities and materials, a site plan with the movement paths, the current process for counting or issuing, and the master data key that the ERP uses. Missing or duplicated asset numbers are the most common blocker and should be cleaned before tagging.
You do. Tag data, middleware configuration, integration source code and as-built documentation are handed over, with a training session for your team and an optional support agreement. There is no dependency on a proprietary cloud unless you choose one.
Not sure whether you need RFID, BLE or UWB?
Send us your asset list, the question you need answered (where is it, how many are there, or when did it move) and a site plan. We reply with a written recommendation, an indicative tag and infrastructure count, and the case for doing nothing where that is the honest answer.
Request a Technology Selection WorkshopRFID Asset Management System
Swedish Technology supplies the complete RFID stack — UHF tags, handheld and fixed readers, gates, antennas, printers and the asset management platform — with the integration and RF engineering behind it.
Request RFID Solution PricingSources & evidence
- GS1 — EPC UHF Gen2 air interface protocol — the passive UHF RFID air interface standard
- GS1 — EPC Tag Data Standard — identifier encoding for RFID and barcode
- GS1 — Barcodes and standards overview
- Bluetooth SIG — Core Specification — includes direction finding (AoA/AoD)
- FiRa Consortium — UWB technology — UWB interoperability and IEEE 802.15.4z profiles
Vendor and product names are trademarks of their respective owners; references are for technical context and do not imply partnership, certification or endorsement unless stated on the vendor's official pages.