From hook engagement safety to end-to-end ladle intelligence. Vision. Physics. AI. Outcomes.
Hook Vision system deployed across 38 ladle bays in 7 plants. Over 18 months of stable live operation.
Ready to scale to Jindal.
Tata solved visibility. Nobody has solved intelligence yet. That's the opportunity.
150 tonnes of liquid steel at 1600°C. One wrong move. Catastrophic consequence.
Hook-trunnion misalignment during ladle pickup is the root cause of most liquid steel spillovers. Ground crew dependency creates exposure in the most hazardous zone.
Crane cabin view of hook engagement is structurally obstructed. Ground signaling staff are essential today — but every second they spend near 1600°C is unacceptable risk.
The crane operator and ground crew have no live superheat data, no ladle lining status, no heat sequencing intelligence. Decisions are made blind — every heat, every shift.
Physics-informed. Operationally sovereign. No cloud dependency required.
Six intelligence modules — all derived from cameras, radar, and physics. No field sensors.
Ladle ID via crane radar + vision. Heat number from MES. Transfer car location, bay assignment — no RFID, no hot-zone tags.
Tap temperature from BOF + transit time → predicted superheat at caster mouth. Physics model. No sensor near heat.
Campaign count, wear model, reline schedule prediction. Reduces reactive reline failures. Cuts refractory cost per campaign.
Vision-confirmed trunnion engagement. No ground crew in hazard zone. Double confirmation: AI detects, crane operator approves.
Live BowTie barrier health — wrong ladle alerts, low superheat warnings, lining life thresholds. GoatSafety AI framework.
Automated pickup authorisation, crane routing, audit log. AI recommends. Crane driver decides. Zero paperwork.
The system acts, logs, and alerts. The operator confirms, not choreographs.
Adjust tap temperature and transit time. See predicted superheat at caster — no field sensor required. Physics tells us what the sensor would say.
From BOF tap data + crane radar transit time
Temperature from tap to caster — predicted every 30 seconds
Indicative outcomes based on Tata Steel deployment data and physics-model validation.
Hook Vision AI eliminates personnel exposure. 100% of ladle bay declared no-man zone.
No manual paperwork. Full audit trail from BOF tap to caster pour. Every heat, every shift.
Vision + Physics model. No sensors near heat. Physics tells us what the sensor would say.
Agentic routing + pre-validated pickups eliminate confirmation delays and mis-sequences.
Wear model vs reactive scheduling. ₹0.8–1.5 Cr refractory saving per campaign (typical).
Before hook engages — not after. AI cross-checks heat sequence before every pickup.
Blast furnace, BOF, continuous casting — domain credibility that pure-tech vendors cannot match. We understand ladle bay operations from shift floor to metallurgical limits.
Thermal models grounded in steel thermodynamics — Stefan-Boltzmann, lining conductivity, superheat decay. Not statistical curve-fitting. Results hold at edge cases and new conditions.
GoatSafety AI barrier framework is the foundation — not bolted on. Every module has a BowTie barrier model. AI recommends. Crane driver decides. Safety stays human-controlled.
Designed for brownfield Indian plants — mixed PLC/DCS environments, variable connectivity, tight CapEx cycles, and the operational realities of Indian steel shift operations.
From first conversation to plant-wide deployment — a phased, low-risk path.
2-day visit to Jindal ladle bay — map current pain points, identify quick wins vs full deployment scope.
Define 1 ladle bay as pilot. Hook Vision AI + Ladle Digital Twin + Superheat Physics Model. Agree baseline KPIs.
8–10 week live deployment. Measure against baseline. Full safety barrier coverage report. No disruption to production.
Plant-wide deployment proposal with phased commercial terms, SLA structure, and training programme.
"When you can see everything, predict what's next, and act before it happens — you don't just run operations. You lead them."