Industrial AI for iron foundries and automotive machining
A plant that casts gray and ductile iron on vertical molding lines and machines the parts afterwards lives with three worlds that rarely talk to each other: the metal — induction furnaces, ferroalloys, nodularisation treatment and the lab —, high-cadence molding and finishing, and a machining shop that runs far slower than the line feeding it. Between those worlds sit paper forms, spreadsheets, closed panels and emails: information that exists but arrives late.
The consequences are concrete and expensive. The master schedule is rebuilt by hand every time a new signal lands. Raw casting stock grows because it is the only cushion against the imbalance between stages. And when a metallurgical question comes up about a brake or chassis part, bounding it by heat number becomes a two-week investigation instead of a two-minute query.
This page collects the twelve canonical IRIS cases applied to this sub-sector: the five Connect capture modes, the two ERP integrations with a financial angle, the traceability marker integration, the three Edge in-line vision archetypes and the agentic case that coordinates all of them against the sector's capital pain. Each one with its real problem, its iLEAN flow and its estimated return range.
Capture without changing habits: photo, voice, email, integrations
The heat sheet at second zero
Payback 4-9 months · from 4-8 h of latency to second zero
See the case →
The induction furnace starts talking
Payback 5-10 months · 3-8% less energy per metric ton
See the case →
The customer alert lands in the schedule
Payback 3-8 months · 30-50% less premium freight
See the case →
Planning while walking the bay
Payback 4-9 months · 45-90 minutes a day recovered
See the case →
First piece released in two taps
Payback 4-8 months · zero unsigned data in the ERP
See the case →
The lab spreadsheet reaches the ERP
Payback 3-7 months · from 4 quality licenses to 1
See the case →
Charge receiving without the queue
Payback 4-8 months · from 45 minutes at the dock to under 10
See the case →
The marker engraves what the order says
Payback 5-10 months · zero mismatched marking on safety parts
See the case →Machine vision on the line, in real time
More cases from this series
- The audit evidence pack, generated as the process happensEvery heat and every changeover leaves evidence organized by standard: IATF 16949, VDA 6.3, CQI-9,…
- Zero stoppages on the customer's line: the three coordinated ringsThree rings — signal, matter and memory — coordinate Connect, Edge and Agents to eliminate customer line…
- The alert that arrives by email, turned into a scheduling decisionPull-ins, delays and incidents arrive by email and messaging. iLEAN cross-checks them against the EDI…
- First-piece release with early human verificationNo AI-captured data enters the ERP unsigned. The line tablet shows what was captured beside its photo…
- The closed induction furnace that finally talksAn external camera digitizes the isolated induction furnace curve: energy-metal-defect correlation with…
- Planning while walking the bay, hands freeThe planner walks. With an industrial earpiece, iLEAN listens, structures and answers in seconds: the…
Book an AI Discovery for your plant: two days with your team to put a number on the casting-to-machining imbalance and prioritize the three cases with the fastest return." ---
We work on your plant's real data, not ours. Assessment with no commitment.
See automotive