Traceability Math

How to Calculate Traceability Coverage, Serialization Workload, and Lot Genealogy Completeness

The core traceability, serialization, and lot genealogy formulas worked end to end, showing exactly where each input comes from and how the numbers resolve.

Traceability coverage is the base metric. Compute it as (units with a complete forward and backward link divided by total units produced) times 100. If you ran 42,000 units in a shift and 40,110 carry both a resolvable parent lot and a captured serial scan, coverage is 40,110 / 42,000 = 95.5 percent. The Traceability Coverage calculator wants three inputs pulled straight from your MES: units produced, units with a valid upstream lot reference, and units with a downstream capture event. Count a unit only if both directions resolve, since a one-sided link fails an audit trace and inflates the number by 3 to 8 points.

Serialization workload converts your unit volume into scan and print events per shift. Workload equals units per shift times serialized levels per unit, divided by throughput per operator-hour. A carton with item, inner pack, and pallet aggregation carries 3 serialized levels. At 30,000 units and 3 levels, that is 90,000 serial events. If a station clears 1,400 verified serials per operator-hour, you need 90,000 / 1,400 = 64.3 operator-hours per shift. The Serialization Workload calculator uses aggregation depth and verified scan rate, not raw belt speed, because verification failures re-queue events and add 5 to 12 percent rework.

Lot genealogy completeness measures how much of the bill of materials resolves to a specific lot at every node. Compute it node by node: completeness equals (BOM component-lot links captured divided by total component-lot links required) times 100. A finished product with 18 components across 3 sub-assemblies needs 18 component-lot links per unit; if 16 resolve, that node is 16 / 18 = 88.9 percent. The Lot Genealogy Completeness calculator aggregates every node into a single tree score, so one missing bulk-ingredient lot at the top can cap the whole genealogy at that node's weakest link.

Recall exposure radius tells you how many downstream units one suspect lot can reach. Exposure equals suspect lot quantity times the fan-out ratio at each stage, chained through the genealogy. If a 2,000 kg resin lot feeds 5 production batches, each batch splits into 4 shipment lots, and each ship-lot averages 620 units, exposure is 5 times 4 times 620 = 12,400 units at risk, not the 2,000 kg you started with. The Recall Exposure Radius calculator multiplies the split factors so you quantify blast radius before an event, which drives how tightly you cap lot sizes.

Scan compliance rate isolates process discipline from system design. It equals (scans that occurred within the defined window and location divided by scans expected) times 100. Expected scans come from routing: a 6-station route with a mandatory scan at 4 stations expects 4 scans per unit. Run 25,000 units and you expect 100,000 scans; if the floor logs 96,500 in-spec, compliance is 96.5 percent. The Scan Compliance Rate calculator separates missed scans from mis-timed and wrong-location scans, because a scan logged at the wrong station still counts against genealogy accuracy even when the count looks right.

The Traceability Gap Score rolls the weak points into one number so you can prioritize. Score each gap as frequency times severity times detectability, similar to an FMEA RPN but bounded 1 to 1,000. A missing pallet-level aggregation that occurs on 4 percent of lots (frequency 6), blocks a full recall trace (severity 9), and is caught only at audit (detectability 8) scores 6 times 9 times 8 = 432. The Traceability Gap Score calculator ranks gaps so a 432 gets closed before a 90, even if the 90 happens ten times more often but resolves in seconds.

Genealogy lookup time is a throughput calculation, not a guess. Average lookup time equals (nodes traversed times seconds per node) plus fixed query overhead. Tracing a unit back through 3 sub-assemblies and 18 component lots traverses roughly 21 nodes; at 0.4 seconds per indexed node plus 3 seconds overhead, that is 21 times 0.4 + 3 = 11.4 seconds. The Genealogy Lookup Time calculator exposes how un-indexed lot tables push per-node time from 0.4 to 6 or more seconds, turning an 11 second trace into a 2 minute one during a live recall clock.

Tie the metrics together with one reconciliation check before you trust any of them. Units produced must equal units with a serial, plus units voided, plus units scrapped, within counting tolerance. If production reports 42,000 but serials plus voids plus scrap sum to 41,300, you have 700 unaccounted units and every downstream percentage is overstated. Set the tolerance at 0.5 percent (210 units here) and investigate anything above it. Feed the reconciled unit count back into the Traceability Coverage and Lot Record Completeness calculators, since a clean denominator is what keeps all five headline numbers defensible in an audit.

Published 2026-07-01.