Centre stage
Keynote
In person & online

From plan to plate: AI that designs out steel waste

  • Date
    9 July 2026
    Timeframe
    14:30 - 15:00 CEST
    Duration
    30 minutes

      In heavy equipment manufacturing, most structural parts are cut from steel plates — and which orders are scheduled together determines how efficiently those parts pack onto each plate. Yet industry practice treats scheduling and cutting as separate decisions: planners batch orders around delivery and capacity, with no visibility into the material waste their choices create downstream.

      This work makes plate utilization a planning-stage objective. The system decomposes each product’s bill of materials to the cut-part level, predicts how a candidate batch will perform on the plate (0.80 correlation with actual utilization), and composes schedules that pack parts efficiently while still meeting delivery and capacity constraints — automatically, and with trigger-based replanning when orders or timelines change.

      In real deployment, plate utilization rose from 80% to 82% under identical demand, with 90%+ of schedules auto-generated and manual effort cut by 60% — an estimated 10M+ RMB in annual material savings. Crucially, this requires no change to cutting equipment or processes: waste is designed out through better decisions alone.

      In an industry where steel is among the world’s largest sources of carbon, every plate saved is embodied carbon avoided. This offers heavy manufacturing a low-cost, scalable path to responsible production and sustainable industry — proving that AI can conserve the planet’s materials, not just optimize output.

      Are you sure you want to remove this speaker?