The operating problem
A Victorian third-party logistics operator adds an interstate service for a bulky-goods customer. Its warehouse system knows what was physically scanned. Its transport system knows what was booked onto the outbound trailer. If the manifest lists eight handling units and the departure scan confirms seven, an order-level status of “dispatched” can hide the short shipment until the receiving depot opens the trailer. That narrows the options to rebooking, a split delivery or a customer escalation.
System landscape
The warehouse owns physical scans, the transport system owns movement instructions and the carrier owns acknowledgements. Dispatch owns authorisation when the two records disagree.
Proposed architecture
Start with one customer, one outbound lane and the receiving depot. Agree which system owns bookings, scans and delivery evidence. Ingest the customer’s despatch advice, warehouse scans, transport manifest revisions and carrier status events through the supported API or agreed file feed. Map the customer order, consignment and handling-unit IDs, retaining the original identifiers. Use the Serial Shipping Container Code (SSCC) when trading partners already supply it. Capture the event time, ingestion time, depot and manifest version so dispatch can distinguish a late feed from absent freight.
- WarehousePallet scans and location
- Transport managementManifest and carrier instruction
- DispatchApprove hold, split or reassignment
- Receiving depotAcknowledged quantity and next action
- Source recordsDespatch advice, handling-unit scans, manifest revisions and carrier events
- ProcessingReconcile expected units against actual scans and show stale feeds
- Decision ownerDispatch supervisor records hold, authorised short-send or reassignment
- Operational outputReceiving depot and customer service get a verified status and next action
System boundaries
The warehouse management system remains authoritative for physical scan events and locations. The transport management system owns the dispatched load, carrier and destination. A carrier delivery event can confirm delivery only for the units it explicitly covers. The integration publishes the agreed status without overwriting source history.
Use existing transport-platform features where they cover the workflow. Build the remaining integration around customer-specific identifiers, milestone mapping and exception ownership. For offline sessions, retain the locally recorded event time and indicate feed freshness at the desk.
What the desk sees before cut-off
Each exception shows the booked unit count, confirmed units, last known scan, feed age, departure deadline and current owner. The dock lead can mark a physical check complete. Only the dispatch role can authorise a short-send or hold. Customer service can see the outcome and supporting evidence with access scoped to its accounts.
At arrival, reconcile the receiving scans against the dispatched quantity. Send unresolved discrepancies to the receiving lead. Attach proof of delivery to the correct shipment leg and keep incomplete evidence visible at billing review.
Reliability and operating controls
Treat a missing scan, a confirmed short, a stale data feed and an authorised split as different exceptions. Route physical checks to the dock lead and departure decisions to dispatch. Preserve an event ledger and deduplicate retries using the source event ID, or the source, unit, event type and timestamp when no event ID is available. Once a manifest is closed, keep a record of late corrections and the person approving them.
Illustrative operating scenario
At 16:55, the Melbourne-to-Sydney manifest expects eight units. Seven have valid load scans and the eighth was moved to a damage bay. The dock lead confirms the location. Dispatch records an approved seven-unit departure and assigns the remaining unit to the next service, with the new ETA and owner recorded. The Sydney depot expects seven. Customer service sees a split shipment and the revised ETA for the eighth unit. A delayed scan from the first trailer is retained as late evidence and sent for review.
Delivery boundaries
The integration makes evidence and holds visible. Dispatch and carrier staff retain load, road and departure decisions. Partner participation depends on the scans and acknowledgements their systems can supply.
Expected business impact
Dispatch could act on an unresolved short before departure. Test time to ownership, late corrections and avoidable follow-up deliveries on the selected lane.
Acceptance test and handover
Replay a representative set of closed loads alongside the warehouse lead and transport planner, covering both normal shifts and cut-off periods. Compare unverified pallets at manifest close, minutes from exception to acknowledged owner, duplicate or false exceptions, customer status corrections and missing delivery evidence. Track the time spent chasing records and follow-on deliveries caused by unresolved shorts. Use the measured volume and avoidable cost across lanes to decide whether rollout is justified.
Before live writes, confirm every blocked or released load can be traced back to its evidence and named owner. Replay duplicate scans, missing carrier updates, corrections after cut-off and partial deliveries. During rollout, publish to dispatch first, then enable outward updates only after sign-off. Give operations a queue owner, feed-health view, replay and reconciliation instructions, and a manual dispatch fallback. Reconcile open exceptions with both depots at each shift change.
Industry evidence
GS1 Australia · Logistic unit identification and despatch advice
Australian National Freight and Supply Chain Strategy
SeaRoad · Warehouse-to-transport integration and proof of delivery