Scattered photos to WMS/TMS and ONVIF?linked claims with SSCC, fact?based settlements

Overview

A parcel carrier’s damage claims lingered because photos, incident notes, and shipment data were scattered across phones, file shares, and separate systems. Evidence arrived late or incomplete, and partners disputed responsibility without a consistent record. Intelligex implemented a claims module that pulled images from warehouse cameras, matched them to shipments and scan events, and enforced service-level timers for carrier responses. Claims moved with complete, time-stamped evidence, negotiations shifted to facts instead of email hunts, and cycle time dropped without extra meetings.

Client Profile

• Industry: Parcel and last-mile delivery
• Company size (range): Multi-terminal network with regional hubs and depots
• Stage: Established Warehouse Management System (WMS), Transportation Management System (TMS), and parcel manifesting tools
• Department owner: Procurement, Supply Chain & Logistics
• Other stakeholders: Claims and risk, Hub and depot operations, Linehaul and last-mile, Customer service, Legal, Finance, IT applications, Key account management, Carrier partners and 3PL handoffs

The Challenge

Damage incidents were documented inconsistently. Dock associates snapped photos on mobile phones, supervisors pulled clips from camera systems when they had time, and incident forms lived in emails or spreadsheets. By the time claims were filed, shipment identifiers did not always match what partners expected. Images lacked context, timestamps were unclear, and the connection between a photo and a specific parcel handoff was weak.

Core platforms were already in place. The WMS tracked inbound and outbound movements, the TMS handled tenders and status, and parcel systems printed labels and submitted manifests. None of these owned a single claims record that combined imagery, scan events, and shipment identifiers with clear timers for response. Partners questioned evidence, internal teams repeated the same data collection steps, and disputes bogged down in back-and-forth emails.

Why It Was Happening

Evidence and identifiers were fragmented. Camera footage sat on local network video recorders; images were exported as ad hoc files without consistent naming; scan events were captured in separate systems. Shipment references varied between tracking numbers, container IDs, and internal load IDs. Without a canonical model tying images, timestamps, and scans to the same parcel, proving when and where damage occurred was slow and error-prone.

Process gaps compounded the issue. There was no formal workflow with timers tied to contract obligations, and responsibilities for gathering evidence were unclear across hubs and depots. Partners received mixed formats and partial records. Reviews focused on reconstructing what happened rather than agreeing on remediation.

The Solution

Intelligex delivered a claims orchestration module that created a single record of each incident. The module pulled footage from dock cameras, extracted stills around key events, and matched them to shipments using label scans and time windows. It assembled parcel identifiers, scan histories, handling notes, and images into a claim package, and started response timers aligned to partner SLAs. Exceptions routed to the right owner with context, and partners received a consistent, auditable bundle. The design leveraged standards for identifiers and timestamps, including Serial Shipping Container Codes (GS1 SSCC), the ONVIF profile for camera access, and ISO 8601 time formats.

• Integrations: Bi-directional sync with WMS for inbound/outbound scans and location; TMS for tenders, status, and route legs; parcel manifesting for labels and tracking numbers; camera systems via ONVIF/RTSP; partner status and updates through APIs defined with the OpenAPI Specification and EDI aligned with X12 transportation messages.
• Canonical claims model: Shipment references (tracking, SSCC, load/stop), scan and handoff events, location and device metadata, images and clips with timestamps, handling notes, and contract references.
• Image and event matching: Time-window extraction of camera frames around scan events; barcode/label matching on stills; confidence flags and operator review when signals conflicted.
• Claim packages: Evidence bundles containing event timelines, images, and handling notes; watermarked previews for sharing; download controls and link expiry for external recipients.
• Timers and workflow: SLA-based timers for partner acknowledgment and response; human-in-the-loop routing to claims analysts, hub ops, or partner managers; reason codes captured for decisions and settlements.
• Validations and guardrails: Checks for missing identifiers, mismatched time zones, partial scan histories, and low-confidence image matches; prompts to reconcile identifiers and re-run extraction before release.
• Dashboards: Aging by claim stage, hotspots by site and route, partner response performance, and repeat issue patterns tied to handling or packaging.
• Permissions and audit: Role-based access for claims, operations, and partners; immutable logs of evidence collection, edits, shares, and approvals.

Implementation

• Discovery: Mapped the incident-to-claim flow across hubs and depots; inventoried camera systems and retention policies; cataloged scan and manifest data sources; reviewed partner claim expectations and SLA terms; identified chronic lanes and sites with unresolved claims.
• Design: Defined the canonical claims and evidence schema; set matching rules for images and scan events; designed SLA timers and escalation paths; established reason codes and a shared glossary for claim statuses.
• Build: Implemented connectors to WMS, TMS, manifesting, and camera systems; built time-window extraction and image matching; created the claim package generator and sharing controls; configured workflow, timers, and dashboards.
• Testing/QA: Replayed past incidents to validate matching and package completeness; tuned time-window and barcode recognition thresholds; ran observe-only mode where claims were assembled alongside the legacy process; enforced human-in-the-loop review for low-confidence matches.
• Rollout: Piloted at select hubs with high claim volumes; kept the email-based path as a fallback; enabled SLA timers and partner sharing after users trusted package quality; expanded by region and partner as exception rates stabilized.
• Training/hand-off: Scenario-based sessions for claims analysts, hub supervisors, and partner managers; quick guides on evidence review and package release; partner briefings on new formats and timelines; transitioned operations to claims and operations teams with IT on call.

Results

Claims moved with complete, consistent evidence. Images, scan events, and shipment identifiers were tied together from the start, and partners saw the same facts the carrier used internally. Response timers prompted timely acknowledgments, and settlements focused on remediation rather than reconstructing basic details.

Operations shifted from cleanup to prevention. Dashboards highlighted hotspots by site, route, and handling step, and supervisors used evidence trends to adjust processes and packaging. Customer service handled fewer escalations, finance closed claims with less back-and-forth, and partner trust in the submission quality improved.

What Changed for the Team

• Before: Photos and notes lived in inboxes and phones; After: Camera frames and images were pulled automatically and tied to scan events in one record.
• Before: Claims lacked clear shipment references; After: Tracking numbers, SSCCs, and load IDs were mapped consistently with timestamps.
• Before: Partners debated partial evidence; After: Claim packages carried timelines, images, and handling notes with clear provenance.
• Before: Follow-ups depended on memory; After: SLA timers drove acknowledgments and escalations with reason-coded actions.
• Before: Root causes were anecdotal; After: Dashboards showed repeat issues by site, route, and packaging for targeted fixes.

Key Takeaways

• Create a single claim record that binds images, scan events, and shipment identifiers so evidence is complete and defensible.
• Pull camera frames programmatically around scan events using standards like ONVIF, and normalize time with ISO 8601 to avoid mismatches.
• Use globally recognized identifiers such as GS1 SSCC on labels to improve matching across systems and partners.
• Enforce SLA-based timers and route exceptions with reason codes so claims progress without manual chasing.
• Integrate with WMS, TMS, and parcel tools; add orchestration and evidence assembly rather than replacing core systems.
• Pilot at high-volume hubs in observe-only mode, tune matching thresholds, then enable partner sharing once packages are consistently trusted.

FAQ

What tools did this integrate with?
The module connected to the WMS for scan events and locations, to the TMS for route legs and status, and to parcel manifesting for label and tracking data. Camera systems were accessed via ONVIF/RTSP, and partner updates flowed through APIs following the OpenAPI Specification or via EDI aligned with X12 transportation messages. Identifiers leveraged GS1 SSCC where available.

How did you handle quality control and governance?
We applied validations for identifier consistency, time-window alignment, and image-match confidence. Low-confidence matches and missing data routed to human-in-the-loop review with reason codes. Claim packages carried watermarked previews and download controls, and every action—evidence capture, edits, shares, approvals—was audit-logged with user and timestamp.

How did you roll this out without disruption?
We ran the module in observe-only mode at a few hubs, assembling claim packages while the legacy email process continued. After users confirmed package completeness and partner feedback was positive, we enabled SLA timers and external shares. The legacy path remained as a fallback during transition and was retired once exception rates stabilized.

How were images matched to specific shipments?
The system used scan events and label data to define time windows around dock activities, then extracted frames from camera streams and matched barcodes or label features where visible. It linked images to tracking numbers, SSCCs, and load IDs, and flagged any ambiguity for human review before release.

What about privacy and data retention for camera footage?
Access to camera streams followed role-based permissions, and only frames relevant to an incident were stored with claims. Timestamps used ISO 8601 for consistency. Retention aligned to policy and contractual obligations, with automatic expiry and audit trails for all views and shares.