Every carrier, sub-assembly and serial on your line reports itself — automatically, without line-of-sight and without an operator stopping to scan. SeekSoft™ MLTS turns physical movement into a timestamped, audit-grade production record your MES and ERP can act on.
System Overview
Production and WIP traceability is the continuous digital record of what was built, where it is, who built it and under what conditions — captured at the moment each event happens rather than reconstructed afterwards. SeekSoft™ MLTS is AOPL's middleware layer for that record. It fuses five identification modalities on one line, filters the noise at the edge, and writes clean, contextualised events into your MES and ERP.
Passive 865–867 MHz tags on carriers, fixtures and parts. Bulk, non-line-of-sight reads at conveyor speed — the backbone of the MLTS event stream.
PRIMARY LAYERDirect part marking and GS1 DataMatrix for piece-part identity where a tag cannot be fitted — laser, dot-peen and label, verified to ISO/IEC 15415 grade.
Presence, orientation and small-component inspection at the station — catching what an identity read cannot: the missing clip, the reversed gasket, the unseated connector.
Torque curves, leak-test results, oven profiles and cycle counters pulled from PLCs and instruments over OPC UA — bound to the serial, not left in a machine-local log.
Badge, biometric or PIN at the station binds the operation to a named, skill-qualified person — so accountability and the skill matrix are enforced, not assumed.
Five capture layers · one event model · one integration contract
The Problem at Scale
Most large manufacturers already have a traceability process. It passes audit on paper and fails in practice — because it depends on people remembering to record things, on labels surviving an environment designed to destroy them, and on data arriving hours after the decision needed it.
Barcode scanning, job cards and move tickets all require an operator to stop, aim and scan. A scan that never happens leaves no trace — the record shows a clean process where a gap actually exists, and nobody finds out until a customer complaint arrives.
Paint-cure ovens run at 180–250 °C, weld cells throw spatter, wash lines flood, and machining floods everything in coolant. Standard PET and paper labels degrade above roughly 85 °C, so identity is routinely lost at exactly the process steps that matter most.
If the finest resolution in your record is a batch or a shift, containment must cover the whole batch or shift. The defect is narrow; the recall is not — and the cost difference between precise and imprecise containment is usually the entire business case.
ERP bookings land at shift end. OEE calculated from those bookings is hours late and structurally wrong: micro-stops, starvation and short stoppages never appear at all, so the bottleneck is identified by opinion rather than by timestamped evidence.
Component lots sit in one system, test results in another, supplier certificates in email, process parameters in a machine log. Nobody holds the unit's full history — it is stitched together by an engineer with spreadsheets, days after it was needed.
Skids, pallets, trolleys, dunnage and returnable containers circulate anonymously. Nobody can say how many are in the loop or where they are, so line stoppages from missing carriers are treated as bad luck rather than a measurable, fixable inventory problem.
Off-line repair is where traceability most often breaks. A unit leaves the sequence, is disassembled, repaired and re-inserted — often at the wrong stage. The repair itself is rarely recorded against the serial, leaving a hole in the very history an investigator will look for.
IATF 16949 §8.5.2 asks you to identify and contain suspect product, with retention often running 10–15 years. Missing or incomplete traceability is among the most common serious audit findings — and preparing the evidence pack consumes weeks of engineering time before every audit.
MLTS Solutions
Fourteen deployable modules on one middleware platform — each with the specific line problem it removes, the evidence it produces, and the plant types it has been engineered for.
WIP Visibility
Tag every carrier, skid, trolley and pallet once, and each becomes a permanent reporting node. Fixed readers at station entry and exit post move events automatically, so WIP location, queue depth and dwell time are live facts rather than shift-end estimates. The carrier tag is reusable for years; only the association to the unit changes.
⚡ Challenges Solved
Quality & Compliance
Bind every serial to its component lots, supplier certificates, process parameters, test results and release decisions as a single queryable record. Trace backward from a field failure to the exact material batch and machine, or forward from a suspect lot to every unit that carries it — in one query, not one engineering week.
⚡ Challenges Solved
Surface & Heat Process
Purpose-built on-metal and high-temperature tags in ceramic, PPS or PEEK housings hold identity through paint-cure cycles at 180–250 °C, e-coat baths, wash lines and vulcanisation. Silicone-free constructions are specified for paint shops, where silicone contamination is an immediate reject.
⚡ Challenges Solved
Line Sequencing
Read the carrier at the sequencing gate and broadcast the build specification downstream before the unit arrives. Each station receives the right variant, torque program and part call-off automatically, and any out-of-sequence body is flagged at the gate rather than discovered at final inspection.
⚡ Challenges Solved
Productivity
Every read is a timestamped event, so station cycle time, dwell, starvation, blockage and changeover duration are measured rather than estimated. Because the data is captured at cycle level, OEE is computed from what the line actually did — not back-calculated from ERP bookings hours later.
⚡ Challenges Solved
Productivity
Read, verify, release. The unit identifies itself at the station, MLTS checks its route and specification, and only then enables the tool, opens the fixture or releases the next operation. A unit that has skipped a step cannot proceed — the control is physical, not procedural.
⚡ Challenges Solved
Productivity
Off-line repair is where most traceability chains break. MLTS keeps the unit identified through removal, disassembly, repair and re-entry, records every repair action against the serial, and enforces the correct re-entry station. Held stock is segregated by system state, not by a paper tag someone might remove.
⚡ Challenges Solved
Productivity
On-metal tags on dies, moulds, fixtures and gauges link every shot, stroke and cycle to the tool that produced it. Validated tool life is counted automatically, calibration expiry blocks use at the station, and changeover kits are verified complete before the line restarts.
⚡ Challenges Solved
Productivity
Tag bins, trolleys, racks and returnable packaging so replenishment is triggered by actual consumption at the line rather than by a card someone has to notice. The same tags close the returnable loop with suppliers and 3PLs, turning unaccounted dunnage into a measured, recoverable asset pool.
⚡ Challenges Solved
Productivity
Bind each operation to the operator who performed it and to their skill matrix. Skill-gated operations release only for qualified people, job-card entry disappears from the operator's workload, and every defect carries an unambiguous station, shift and person attribution for the 8D.
⚡ Challenges Solved
Productivity
Verify that the right engine, battery pack, harness, gearbox or module married to the right parent unit — at the moment of marriage, not at end-of-line test. Kits are confirmed complete at the pick, and the sub-assembly's own genealogy is folded into the parent's record rather than kept in a separate system.
⚡ Challenges Solved
Productivity
Capture torque curves, leak-test and e-test results, oven profiles and consumable lot numbers against the serial as the operation completes. Adhesive, sealant, solder paste and lubricant lots are bound to the units they went into, and shelf-life-expired consumables are blocked at the station.
⚡ Challenges Solved
Productivity
Portal readers at the finished-goods store and dock verify the physical load against the ASN before the vehicle leaves. FIFO is enforced by system state rather than by stacking convention, and the dispatch event becomes the shipping evidence that settles a claim months later.
⚡ Challenges Solved
Productivity
Answer "which units are affected?" in seconds, with a defensible evidence pack behind it. MLTS holds the traceability map, retention schedule and change-control history that IATF surveillance and OEM customer audits ask for — including the recall-scoping demonstration auditors increasingly want run live.
⚡ Challenges Solved
Five layers, one event model. Identity is applied once, captured automatically, given meaning at the edge, written into your enterprise systems, and returned to the line as a decision.
Tags · DPM · Labels
UHF tags on carriers and fixtures, DataMatrix on piece parts, chosen and validated for the actual process environment.
Readers · Vision · PLC
Fixed readers, portals, handhelds, cameras and OPC UA tags collect events without stopping the operator or the line.
SeekSoft™ Edge
Stray reads filtered, direction resolved, events bound to work order, station, operator and shift — at the edge, not in the cloud.
MES · ERP · QMS
ISA-95 aligned interfaces into SAP, Oracle and your MES — IDoc, BAPI, OData, REST, MQTT and B2MML, with store-and-forward.
Andon · Analytics
Live andon at the station, alerts to supervisors, and genealogy queries that scope a containment in seconds.
Business Impact
RFID traceability is not a reporting upgrade. It changes who does the work, when the data exists, and what decisions become possible — across production, quality, maintenance, logistics and finance.
| Business process | Manual / barcode today | With SeekSoft™ MLTS |
|---|---|---|
| WIP visibility | Known at shift-end booking; position inferred from the last scan | Live position, dwell and queue depth per station, updated on every read |
| Production reporting | Operator keys job cards; data entry competes with cycle time | Reporting is a by-product of movement; operator does no data entry |
| Recall containment | Scoped to a batch or a shift; days of manual reconstruction | Scoped to affected serials; query answers in seconds with evidence pack |
| OEE and downtime | Back-calculated from ERP bookings; micro-stops invisible | Cycle-level measurement; micro-stops, starvation and blockage separated |
| Quality investigation | Engineer stitches lots, tests and logs across four systems | One genealogy record per serial, queryable forward and backward |
| Audit preparation | Weeks of evidence assembly before each IATF or OEM audit | Traceability map, retention and change history maintained continuously |
| Line-side inventory | Buffer stock sized for uncertainty; carriers unaccounted | Consumption-triggered replenishment; carrier pool measured and recovered |
| Tooling and maintenance | Counts kept in diaries; life and calibration tracked manually | Automatic stroke and cycle counting; expiry blocks use at the station |
Siemens' True Cost of Downtime research puts unplanned downtime at roughly 11% of revenue across the Fortune Global 500 — about US$1.4 trillion a year. Live cycle-level data is how you find the share of it that is schedulable.
Published OEE studies attribute more of the decline to changeover than to any other single cause. Timestamped carrier reads measure changeover duration exactly, turning the biggest loss into the most improvable one.
Rework and scrap from outdated specifications and process errors run at an estimated 2–5% of annual revenue. Station interlocks stop the wrong operation before it is performed rather than detecting it later.
Human error accounts for around a quarter of stoppages. Removing manual scanning, manual job cards and manual sequence boards removes the largest surface on which that error occurs.
Automotive retention is commonly aligned to vehicle service life. A digital genealogy record makes that obligation a storage decision rather than a warehouse of paper nobody can search under pressure.
EV and industrial batteries above 2 kWh placed on the EU market must carry a Digital Product Passport from 18 February 2027. Serial-level production data is the input; plants without it cannot generate one.
Figures cited from published industry research (Siemens True Cost of Downtime; OEE loss studies; manufacturing documentation benchmarks) and from EU Regulation 2023/1542. They are sector benchmarks for sizing a business case, not AOPL project results — your own baseline should be measured during the line survey.
Line-qualified hardware selected against the actual process — temperature, chemistry, metal, vibration and read geometry — then commissioned, RF-surveyed and handed over with a spares catalogue.
4 and 8-port UHF readers for station and conveyor read points, with PoE and on-reader filtering.
Conveyor tunnels and dock portals that verify full pallets and cartons at line and vehicle speed.
Ceramic, PPS and PEEK tags rated to 250 °C, IP68/IP69K, silicone-free options for paint shops.
Industrial print-and-encode for smart labels, with encode verification and reject handling on the line.
Rugged Android UHF handhelds for cycle counts, audits, line-side verification and rework tracking.
Line-side screens and stack lights driven by MLTS events — takt, target versus actual, and hold status.
Circular and linear polarised panels, near-field and shelf antennas, selected by read-zone geometry.
DIN-rail edge compute running read filtering, business rules and store-and-forward when the network drops.
Our traceability engineers will walk your line, count the read points, test tag performance against your metal, heat and fluids, and return a station-by-station design with a costed business case.
Everything engineering, quality and IT teams ask before committing a running line to RFID traceability.
Production and WIP traceability with RFID is the practice of giving every unit, carrier, pallet or tool a unique electronic identity and reading it automatically at defined points along the line. Each read writes a timestamped event — what was at which station, when, for how long, with which operator, tooling, component lot and test result. The accumulated events form a genealogy record per serial number, so a plant can answer forward questions (where is this order now?) and backward questions (what went into this unit?) without manual reconstruction.
A barcode requires line of sight and a deliberate human action; RFID does not. A gantry or portal reader captures tags in bulk, at speed, through dust, oil, paint mist and packaging, with no operator step in the cycle. Practically, that changes what data you get: barcode gives you the scans someone remembered to perform, RFID gives you every movement. Most plants keep barcode where a single deliberate scan is already reliable and adopt RFID where scanning competes with takt time.
Yes, with the right tag class and antenna design. On-metal UHF tags use a ground plane and spacer to work when mounted directly on steel or aluminium. High-temperature tags survive e-coat and paint bake cycles in the 200–230 °C range for repeated passes, and ceramic variants go higher. Welding and VFD-heavy areas generate RF noise, which is handled through antenna placement, polarisation, shielding and read-zone tuning. This is exactly what the site survey establishes before any commitment on read rates.
MLTS sits between the read layer and your systems of record. Northbound it publishes to SAP S/4HANA and ECC, Oracle, Odoo, Ignition, Wonderware and custom MES through REST, OData, IDoc, BAPI, MQTT and database interfaces. Southbound it speaks to PLCs and station controllers over OPC UA, Modbus TCP and Ethernet/IP for interlocks and andon signalling. It does not replace your MES or ERP — it removes the manual data entry that currently feeds them.
Part genealogy is the as-built record linking a finished serial number to every component lot, sub-assembly serial, process parameter, test result, tool and operator that contributed to it. OEMs and regulators ask for it because it converts a recall from a batch-wide event into a serial-scoped one. Under IATF 16949 and most customer-specific requirements, a supplier must demonstrate the ability to isolate suspect product; genealogy is the evidence that makes that isolation defensible and fast.
A single-line pilot typically runs 8–12 weeks from survey to steady state: two weeks of RF site survey and tag qualification, three to four weeks of hardware installation scheduled around planned shutdowns, two to three weeks of integration and rule configuration, then a parallel-run period where RFID data is validated against existing bookings before the manual process is retired. Plant-wide rollouts extend the installation window but reuse the same tag, antenna and integration design.
Cost is driven by read-point count, tag class and tag reuse rather than by line length. Returnable carriers and pallets are read thousands of times, so tag cost amortises to near zero; per-unit tagging on high-value assemblies carries a real consumable cost that must be justified. ROI is built from four measurable lines: recovered production hours, reduced rework and scrap, released working capital from smaller WIP buffers, and engineering hours no longer spent assembling audit and warranty evidence. We size all four against your measured baseline, not against benchmarks.
Reading continues. Edge gateways hold the read filtering and business rules locally, buffer events in a local store when upstream connectivity drops, and forward them in order when it returns. Station interlocks and andon signalling keep operating on cached rules, so a WAN outage degrades reporting latency rather than stopping the line. For air-gapped or restricted-network plants, MLTS runs fully on-premise with no external dependency.