Real-time material movement across an integrated steel campus: 40% faster than the global benchmark
An RFID + SAP S/4HANA system handling ₹1,20,000 crore of inventory across a multi-million-ton steel campus: and outrunning the global vendor it replaced.
Outcomes
40%
faster
Than the prior global benchmark vendor
₹1,20,000 Cr
inventory
Under live tracking
< 5
min
Movement-to-ERP latency
99.95%
uptime
Including 72hr WAN ride-through
20–30%
reduction
In demurrage cost
The challenge
- Material movement across 800+ acres tracked on paper challans — reconciliation took 5–7 days.
- Global MES vendor’s system had not been able to keep up with the velocity of the integrated campus.
- SAP MM/EWM reconciliation against actual yard state was a daily fire-fight.
- Demurrage and rake-detention costs running into crores per month.
What we built
- UHF RFID at every gate, every internal weigh-bridge and every dispatch zone.
- Custom event-driven middleware handling ~120M reads per month with a settled-event model.
- Real-time SAP S/4HANA bridge — movement to ERP latency < 5 minutes.
- Yard, port and rake tracking with dispatch-sequence optimisation.
- Edge-buffered to ride through 72 hours of WAN outages without dropping events.
The scale problem
An integrated steel campus across hundreds of acres, moving raw materials, intermediates and finished products at a velocity that no manual or barcode-based system could keep up with. The prior MES vendor — a recognised global brand — had been incrementally upgraded for years and was now bottlenecked at the architectural level.
The architectural shift
We replaced the polling-based model with an event-driven stream architecture. Every read at every reader, every weigh-bridge tip, every gate event became a durable event in a Kafka backbone. Settlement logic ran as stateless services that produced canonical movement events. SAP got a clean stream of business-meaningful postings, not a firehose of raw reads.
The benchmarking
Run on the same hardware as the prior vendor, our stack handled 40% more throughput at lower CPU utilisation. The benchmarking exercise was done by the customer’s own engineering team. That number is what unlocked the multi-plant rollout.
Tech stack
- • Impinj enterprise readers
- • Custom edge agents
- • Kafka stream layer
- • SAP S/4HANA (IDoc/BAPI/OData)
- • TimescaleDB + PostgreSQL
Timeline
9-month phased rollout across the integrated campus, plant-by-plant.
Related field guides
RFID & Industrial Traceability
RFID & Industrial Traceability in India: The 2026 Field Guide
A practitioner’s field guide to industrial RFID in India — frequencies, tags, readers, software architecture, and the patterns that survived 100M+ tag reads.
MES & Smart Manufacturing
Manufacturing Execution Systems (MES) in India: The 2026 Practitioner’s Guide
A field guide to choosing, implementing and operating a Manufacturing Execution System in Indian plants — written from 50+ deployments across process and discrete industries.
RFID & Industrial Traceability
RFID + SAP Integration: The Patterns That Work
A clean SAP integration is the difference between RFID that runs the warehouse and RFID that crashes the warehouse. The patterns we use across deployments.
Talk to the team
Want a reference call with this customer?
Under NDA, we facilitate direct conversations with selected reference clients.
Request a reference call