How do I set up an asset tracking platform for chain-of-custody and serial number traceability through manufacturing?

To set up an asset tracking platform for chain-of-custody and serial number traceability through manufacturing, think in terms of people + process + data model + capture points + controls.

1) Define the traceability scope

Start by deciding exactly what must be tracked:

• Serialized finished goods: each unit gets a unique serial number
• Subassemblies/components: if you need genealogy
• Lots/batches: for raw materials or bulk items
• Tooling/fixtures: if they affect quality or custody
• Locations and custody: warehouse, line, WIP, QA, shipping, customer
• Events: receipt, issue to line, assembly, test, pack, ship, return, scrap, rework

If you need chain-of-custody, every movement or status change should be an auditable event with:

• who
• what
• when
• where
• why
• from/to state

2) Design the core data model

At minimum, your system should have these objects:

• Item/Asset Master

  • SKU / part number
  • description
  • serial number or lot number
  • revision / version
  • lifecycle status

• Identity/Genealogy

  • parent-child relationships
  • serial-to-lot associations
  • BOM linkage
  • “used in” / “built into” relationships

• Custody Events

  • event type: receive, move, pick, issue, consume, assemble, inspect, pack, ship, return, quarantine, scrap
  • timestamp
  • user/system
  • source location
  • destination location
  • reference document (work order, PO, shipment, NCR)

• Locations

  • site, building, warehouse, line, cell, bin, truck, customer
  • optional GPS/geofence if field assets matter

• Actors

  • employee, contractor, machine, system integration, customer

• Documents

  • work orders
  • traveler/route cards
  • quality records
  • shipping docs
  • certificates / test results

3) Decide where serial numbers are created

You need a rule for serialization authority:

• Generated at ERP order creation
• Generated at MES work order release
• Generated at pack-out
• Pre-assigned by supplier
• Generated centrally by the asset platform

Best practice:

• Make the serial number globally unique
• Use a controlled format with a prefix, plant code, date/sequence, or check digit
• Prevent re-use forever, even for scrapped units

4) Build capture points into the manufacturing flow

Traceability fails when data is entered too late or manually. Add capture points at every handoff:

Common capture points

• Receiving dock: scan PO, lot, serial
• Incoming QA: inspect and accept/reject
• Warehouse issue: pick to WO
• Line side: issue to operation
• Assembly: consume components into parent serial
• Test station: record test results against serial
• Pack-out: assign packaging serials/case IDs
• Shipping: confirm shipment and custodian transfer
• Returns/service: reopen custody chain

Recommended technologies

• Barcode/QR: cheapest and most common
• RFID: good for high-volume or non-line-of-sight
• Mobile scanners/tablets
• Stationary scanners at gates/conveyors
• Integration with test equipment / PLCs / machine vision if you need automatic capture

5) Use event-driven traceability

Instead of only storing “current status,” store a full event history. This is essential for chain-of-custody.

Example event sequence:

1. Serial created
2. Received into warehouse
3. Issued to work order
4. Consumed into parent serial
5. Tested
6. Packed
7. Shipped

This lets you answer:

• Where was item X at any time?
• Who handled it?
• What parts went into serial Y?
• Which shipped units contained defective lot Z?

6) Implement genealogy

For manufacturing traceability, genealogy is often the most important feature.

You want to know:

• Which component serials/lot numbers went into each finished good serial
• Which work order built it
• Which operator/machine/test station touched it
• Which revision/BOM was used
• Which downstream customers received affected units

A good genealogy model supports:

• one-to-one
• one-to-many
• many-to-one
• many-to-many relationships

7) Enforce custody rules and permissions

Define who can do what:

• Warehouse can receive/move/issue
• Operators can consume/assemble
• QA can hold/release/quarantine
• Shipping can transfer custody
• Admins can correct only with reason codes

Add controls for:

• role-based access
• approval workflows
• dual sign-off for critical changes
• reason codes for overrides
• immutable audit trail

8) Integrate with ERP, MES, WMS, QMS

Don’t build a silo. Most successful platforms integrate with:

• ERP: item master, sales orders, procurement, inventory valuation
• MES: work orders, operations, production events
• WMS: location/bin inventory, picking, receiving
• QMS: inspections, NCRs, CAPAs, holds/releases
• PLM: engineering revisions, BOM changes
• Shipping/TMS: shipment confirmation, carrier tracking

Integration patterns:

• API-first
• event streaming/webhooks
• scheduled sync for master data
• real-time transaction posting for custody changes

9) Decide on data integrity and compliance requirements

If you operate in regulated industries, you may need:

• FDA 21 CFR Part 11
• ISO 9001 / ISO 13485
• aerospace/defense traceability standards
• audit logs that are tamper-evident
• electronic signatures
• timestamp accuracy
• record retention policies

Use:

• append-only audit logs
• immutable event storage or hashing
• backup and disaster recovery
• strict time synchronization
• validated workflows for regulated environments

10) Build exception handling

Real factories have exceptions. Your platform should support:

• damaged items
• lost serials
• partial assemblies
• rework
• scrap
• quarantine/hold
• substitution approvals
• returns from customer
• stolen/missing assets
• duplicate serial detection

Every exception should create an auditable event with reason and approver.

11) Create reports and traceability queries

You’ll want standard queries like:

• Full forward genealogy: component → finished goods → shipped customers
• Full backward trace: customer serial → all component lots and process steps
• Chain-of-custody report: every handoff with timestamps and users
• Recall impact analysis: which lots/serials are affected
• WIP status and aging
• Inventory by location/status

12) Choose the system architecture

Typical architecture:

• Frontend: web app + mobile scanning app
• Backend: API services for assets, inventory, genealogy, events
• Database: relational DB for master data + event log; optionally event store
• Integration layer: ERP/MES/WMS connectors
• Identity/access: SSO, RBAC, audit
• Hardware: barcode printers, scanners, RFID readers

A practical pattern is:

• transactional relational database
• immutable event table
• service layer enforcing business rules
• analytics/reporting warehouse for queries

13) Pilot before scaling

Start with one line, one site, or one product family.

Pilot goals:

• verify scan points are realistic
• confirm serial format and label durability
• test genealogy capture
• measure cycle-time impact
• train operators
• fix edge cases before rollout

Success metrics:

• scan compliance rate
• traceability completeness
• time to locate asset
• time to perform recall analysis
• data error rate
• audit findings

14) Practical implementation checklist

Here’s a good rollout sequence:

1. Define traceability requirements and regulatory constraints
2. Standardize item/serial/lot/location schemas
3. Decide event types and custody rules
4. Map manufacturing workflow and scan points
5. Integrate with ERP/MES/WMS/QMS
6. Deploy printers/scanners/mobile devices
7. Create audit logs and permissions
8. Pilot on one product line
9. Validate reports and recall queries
10. Scale to additional lines/sites

15) What “good” looks like

A strong asset tracking platform should let you answer, in seconds:

• Where is this serial number now?
• Who had custody of it at every step?
• What parts, lots, and processes went into it?
• Was it inspected/tested, and by whom?
• Which customers received affected units?
• Can we prove the record history is complete and unaltered?

If you want, I can also provide:

1. a reference architecture diagram,
2. a database schema for chain-of-custody and genealogy, or
3. a step-by-step implementation plan for a specific factory type.