LogiTrack — RFID & AI-powered traceability for warehouse, production and shipping

The challenge

Modern logistics operations face a compounding problem: goods move faster than information. Manual barcode scanning cannot keep up with high-throughput warehouses, multi-stage production lines, and strict regulatory traceability requirements. Shipping errors, inventory blind spots, and counterfeiting risks erode margins and customer trust.

LogiTrack was designed from the ground up to eliminate these gaps — replacing line-of-sight scanning with bulk RFID reads, manual checks with AI-powered validation, and disconnected spreadsheets with a single, real-time data layer spanning every operational touchpoint.

Cloud-native, edge-ready architecture

LogiTrack runs on a containerised stack (Docker/Kubernetes) with serverless API functions via Apache OpenWhisk, enterprise identity management through Keycloak, and data virtualisation via Dremio feeding Power BI dashboards. Edge computing support means RFID gates and industrial label printers operate locally at critical operational points, with data synchronisation to the central platform.

Two deployment modes are available. The integrated configuration with eGO CRM APPs on the Power Platform provides centralised governance, low-code automation via Power Automate, self-service portals via Power Pages, and Copilot Studio virtual assistants. The standalone mode offers full autonomy for air-gapped environments, extreme customisation needs, or organisations without Microsoft 365 licences. Both share the same core engine.

Warehouse management

A rugged, offline-first mobile app handles every warehouse flow: goods receipt with ASN validation, intelligent putaway with optimal location suggestions, RFID-guided picking, cyclic and mass inventories without stopping operations, and automatically tracked internal transfers. The touch-optimised interface works with gloves, in low light, and supports multi-language, multi-site deployments.

Production traceability

Every production lot receives a unique EPC-encoded RFID tag. As it moves through work phases, gates automatically register transitions and update the work order status in real time. Machine addressability via RFID auto-configures processing parameters for each lot. Material consumption is captured via the mobile app with automatic backflush. Cycle-time KPIs and OEE are calculated live for every work centre and operator.

Zero-touch shipping validation

After packing, each handling unit (HU) is sealed with an RFID label. As it passes through the exit gate, the system automatically validates contents against the shipping order — verifying quantities, articles, and destination. A perfect match triggers automatic SAP synchronisation and transport-document generation. Any discrepancy blocks the shipment with an immediate operator alert. Result: 99.5% reduction in shipping errors.

Returns, refurbishment and second channel

Returned goods enter automatic virtual quarantine. The system reads the RFID tag, recovers the product's full history, and suggests the next action — refurbish, remarket via second channel, or scrap — based on configurable rules. Refurbishment triggers auto-generated work orders. Once completed, the product is re-certified with a blockchain-anchored state update for full transparency.

External processing at third-party workshops is tracked end-to-end: outbound registration, dwell-time monitoring with automatic delay alerts, inbound RFID verification to prevent fraudulent substitution, and real-time visibility during every external phase.

NFC consumer experience and authenticity

End customers tap their smartphone on the NFC-enabled product to open a white-label brand app showing the complete product journey — from raw materials to final delivery, with timestamped, geolocated records for every phase. A blockchain-backed certificate of authenticity prevents counterfeiting. Customers can register for warranty extension and personalised support, while brands gain post-sale behavioural insights.

AI and intelligent automation

LogiTrack embeds a full AI ecosystem: Weaviate for vector indexing, Elysia for natural-language processing, and n8n for orchestrating intelligent pipelines. Key capabilities include:

• Semantic search — natural-language queries across ASNs, delivery notes, and transport documents, returning relevant results even from imprecise or partial queries.
• RFID–video correlation — ONVIF camera integration to automatically link RFID events with video recordings for audit, discrepancy investigation, and flow optimisation.
• Anomaly detection — automatic identification of irregular patterns in gate throughput, scrap rates, cycle times, and operator behaviour, with proactive alerts before issues impact operations.
• Conversational AI — chat interfaces integrated with Microsoft Teams, brand portals, and customer-facing bots, enabling operators to query the system in natural language for reports, decisions, and support.
• Narrative reporting — the AI engine auto-generates natural-language reports explaining trends, highlighting anomalies, and suggesting corrective actions. Managers receive personalised executive summaries instead of interpreting raw dashboards.

Enterprise-grade security and compliance

Identity and access management via Keycloak supports SSO (SAML, OAuth2), MFA (OTP, biometrics, smart cards), Active Directory integration, and granular RBAC/ABAC authorisation. All API traffic passes through APISIX with JWT authentication, mutual TLS for critical connections, and rate limiting. Data is encrypted AES-256 at rest and TLS 1.3 in transit.

LogiTrack is designed for native compliance with EPCglobal/GS1, ISO 18000-63 (RFID UHF), ISO 14443/15693 (NFC), and ONVIF. GDPR is addressed through privacy-by-design, right to erasure, and immutable audit trails. The AI components are classified as low-risk under the EU AI Act, with full explainability and transparency.

An integrated SIEM provides continuous security monitoring with multi-level event correlation, automatic anomaly alerting, documented incident-response procedures, and encrypted backups with configurable retention and guaranteed RTO/RPO.

Monitoring and analytics

Prometheus and Grafana deliver real-time metrics on gate throughput, RFID read errors, API latency, and infrastructure health. Elasticsearch centralises application and audit logs. Dremio virtualises heterogeneous datasets — RFID events, SAP orders, production work orders, warehouse movements — into a single semantic data layer consumed by Power BI for interactive dashboards.

Dashboards are role-segmented: executives see high-level trends, cross-site comparisons, and aggregated KPIs; operations managers get real-time, granular views per gate, warehouse zone, production line, and operator. Advanced features include interactive gate heatmaps, event timelines with drill-down, RFID–video correlations, configurable threshold alerts, and data export for external analysis and compliance.

Results

• 99.5% reduction in shipping errors through automatic RFID gate validation against planned orders.
• 20× faster inventories — mass counts completed in minutes instead of hours, with zero operational downtime.
• Complete product biography — every item tracked from inbound receipt through production, shipping, returns, and end-consumer authentication.
• AI-driven decision support — anomaly detection, predictive analytics, and narrative reporting replace manual analysis.
• 8–12 month ROI payback through error reduction, time optimisation, and productivity gains.
• Blockchain-certified authenticity for premium products, accessible to end consumers via NFC tap.

Technologies used