Forgiron ships as six products mapped to the forge, not to abstract capabilities. Heatrix, Forgeon and Diesmith control a process step, Flawnex is the shared inspection and quality backbone, Forgebot is the hands, and Forgetwin is the twin and planner over the whole line. Start with one on the Cell tier and expand as each proves its number.
Illustrative scenario: Cell F-2 · 2,500 t closed-die press · 42CrMo4 connecting rod, one lot, one approval gate
Heatrix controls billet heating and heat-treat, the thermal state that sets everything downstream. Forgeon controls the stroke, fill and grain flow. Diesmith manages die life, lubrication and wear. Flawnex senses laps, folds, underfill, cracks and internal defects and models properties. Forgebot handles hot billets, forgings and dies. Forgetwin simulates the whole forge and sequences yield and takt, handing proven setpoints back.
Heatrix senses billet temperature, core-to-surface spread, soak and furnace or induction state from thermal imaging, pyrometry and coil or burner telemetry, predicts the downstream effect of the thermal state on fill, grain flow and properties, and continuously adapts induction power, furnace zones and soak time to hit target temperature at minimum energy.
Forgeon senses press or hammer force, stroke, energy, billet position and die fill in real time, predicts fill, grain flow and defect risk from the incoming thermal state supplied by Heatrix, and adapts stroke schedule, blow energy and speed to hit target fill and grain flow at minimum blows and energy.
Diesmith senses die temperature, wear, heat-checking and lubricant coverage with thermal and vision, predicts remaining die life and thermal-fatigue risk, and continuously controls lubrication dosing, spray and die cooling to minimise friction, sticking and wear while hitting fill.
Flawnex fuses vision, ultrasonic, MPI, thermal and dimensional data to detect and predict laps, folds, underfill, cracks and internal defects, and models grain flow, hardness and mechanical properties against spec.
Forgebot drives robotic and manipulator handling of hot billets, forgings, dies and trim or inspection parts, perceiving part pose, temperature and grip, planning safe paths around presses and furnaces, and coordinating takt with Heatrix, Forgeon and Flawnex so parts arrive at the right station at the right temperature and time.
Forgetwin simulates billet heating, die thermal state, press stroke, fill, stress, die load, trim and heat-treat, as-forged against as-designed, to hit target soundness, dimension and properties before the stroke, then sequences and balances the moving line, die allocation, heating batches, press and heat-treat queues and robot takt, for maximum yield at minimum scrap, die wear and energy.
Autonomous billet heating and heat-treat: the exact forging temperature, every heat.
Every forging starts with heat. Induction and furnace billet heating must hit exact forging temperature and soak without overheating, decarburization, cold shuts or uneven core-to-surface spreads, and heat-treat must land exact hardness and properties. Today this runs on fixed setpoints, thermocouple spot checks and retiring furnace-operator craft. A heating drift or a heat-treat excursion silently seeds underfill, cracks, poor grain flow and wrong hardness, discovered only after the part is forged, quenched or scrapped, while furnaces burn the forge's single largest energy bill.
Heatrix senses billet temperature, core-to-surface spread, soak and furnace or induction state from thermal imaging, pyrometry and coil or burner telemetry, predicts the downstream effect of the thermal state on fill, grain flow and properties, and continuously adapts induction power, furnace zones and soak time to hit target temperature at minimum energy. It carries the same closed loop into heat-treat, austenitize, quench and temper, to hit hardness and property targets, under human-approved graduated autonomy.
Automotive and driveline, aerospace and defense, energy (turbine, oil and gas, nuclear) and industrial forging makers running induction and furnace heating plus heat-treat lines; induction, furnace and heat-treat equipment OEMs.
Named in the product brief as the design
Planned
Training thermal-spread and property surrogates plus physics-informed thermal FEM over high-rate thermal video and process telemetry, and low-latency edge inference for closed-loop heating, are GPU-bound. CPU-only cannot meet the training throughput or the sub-second control latency.
Advisory temperature and soak optimisation on one induction or furnace line using existing pyrometry and thermal cameras; graduate to closed-loop power and zone control under operator approval, then heat-treat. First metric: energy per heat or thermal-defect scrap on a named cell.
Tier. Cell tier, $15,000 per heating or heat-treat cell per month, expanding to Plant and Enterprise.
Open questions
Autonomous press and hammer control: fill and grain flow decided before the stroke.
The stroke is the moment of truth. Presses and hammers, open-die, closed-die, ring-rolling and upsetting, must deliver exact force, energy and die fill to shape the part and align grain flow for fatigue strength without laps, folds, underfill or cracks. Today stroke schedules are fixed, tuning is craft-bound and success is judged only after the part cools. A wrong blow energy, a mis-timed stroke or an off die fill wastes a heated billet, forces rework or scrap and quietly degrades grain flow on safety-critical forgings.
Forgeon senses press or hammer force, stroke, energy, billet position and die fill in real time, predicts fill, grain flow and defect risk from the incoming thermal state supplied by Heatrix, and adapts stroke schedule, blow energy and speed to hit target fill and grain flow at minimum blows and energy. Every schedule is validated against the Forgetwin before setpoints go live, and press control stays supervised by engineer approval until the part family is qualified.
Closed- and open-die forging, ring-rolling and upsetting shops in automotive and driveline, aerospace and defense, energy and industrial forging; press, hammer and ring-mill OEMs.
Named in the product brief as the design
Planned
Training metal-forming FEM surrogates and grain-flow models over large multivariate press and vision telemetry, plus real-time sub-100 ms edge inference in the stroke loop, is GPU-bound. CPU-only cannot meet the training throughput or the stroke-cycle control latency.
Advisory stroke-schedule and fill optimisation on one press using existing force and stroke sensors and cameras; graduate to bounded closed-loop stroke control under operator approval. First metric: scrap, rework or energy per forging on a named press.
Tier. Cell tier, $15,000 per press, hammer or ring-mill cell per month, expanding to Plant and Enterprise.
Open questions
Autonomous die management and lubrication: more forgings per die, fewer die changes.
Dies are the most expensive, most defect-sensitive assets in the forge. They wear, heat-check and thermally fatigue under repeated high-force, high-temperature strokes, and lubrication and cooling must be dosed exactly to control friction, fill, sticking and die life. Today die spray and cooling run on fixed recipes, die condition is judged by eye, and a die is pulled reactively, after wear, heat-checks or a die-fill defect already scrapped parts. Unplanned die changes are a top driver of forge downtime, and worn dies quietly seed underfill, laps and dimensional drift.
Diesmith senses die temperature, wear, heat-checking and lubricant coverage with thermal and vision, predicts remaining die life and thermal-fatigue risk, and continuously controls lubrication dosing, spray and die cooling to minimise friction, sticking and wear while hitting fill. It schedules die changes and refurbishment before they cause scrap, under human-approved graduated autonomy.
Closed- and open-die forging, ring-rolling and upsetting shops; die-engineering groups; die-shop and lubricant or spray-equipment OEMs across automotive, aerospace, energy and industrial forging.
Named in the product brief as the design
Planned
Training die-stress FEM surrogates and multimodal wear and heat-check models over high-rate thermal and vision data, plus low-latency edge inference for between-stroke spray control, is GPU-bound. CPU-only cannot meet training throughput or the spray-cycle control latency.
Advisory die-life prognostics and spray optimisation on one cell using die thermal cameras and press data; graduate to closed-loop spray and cooling control and predictive die-change scheduling. First metric: forgings per die or unplanned die changes on a named cell.
Tier. Cell tier, $15,000 per forging cell per month as the die module, expanding to Plant and Enterprise.
Open questions
See every lap, fold, crack and internal defect, and the grain flow behind it.
Forgings are the most safety-critical metal parts in the world, and their defects are the hardest to see: laps, folds, underfill, cold shuts and surface cracks on the outside; internal cracks, inclusions and grain-flow faults on the inside; plus hardness and dimensional deviations. Inspection today is ultrasonic, MPI, vision and dimensional, done as slow, spot, often manual checks by a scarce, retiring NDT and metallurgy workforce. A missed lap or internal crack becomes a downgrade, a quarantine, a requalification or a safety-critical field escape.
Flawnex fuses vision, ultrasonic, MPI, thermal and dimensional data to detect and predict laps, folds, underfill, cracks and internal defects, and models grain flow, hardness and mechanical properties against spec. It couples every detection back to the upstream heat, stroke and die state that caused it, dispositions parts against AS9100, IATF 16949, NADCAP and ASTM criteria with cited evidence, and moves inspection toward 100 percent coverage under human-approved review.
Aerospace and defense, energy, automotive and driveline and industrial forging makers with safety-critical qualification; NDT, ultrasonic, MPI and machine-vision equipment OEMs.
Named in the product brief as the design
Planned
Training multimodal defect and property models over large image, ultrasonic and MPI datasets, generating synthetic rare defects, and real-time sub-100 ms edge inference for full-coverage inspection are all GPU-bound. CPU-only cannot meet training throughput, synthetic-data generation or inline inspection latency.
Real-time surface lap, fold and crack vision detection on one line using existing cameras; add ultrasonic and MPI fusion and standards-grounded disposition; graduate toward full-coverage inspection autonomy. First metric: inspection escapes, scrap or downgrade on a named line.
Tier. Cell tier, $15,000 per inspection cell per month, bundled in Plant and Enterprise.
Open questions
Autonomous handling of hot billets, forgings and dies: off the human, onto the robot.
The forge is one of the most brutal, dangerous and rate-limiting workplaces in manufacturing. White-hot billets and forgings weighing kilos to tons must be moved from furnace to press to trim to heat-treat to inspection, and heavy dies must be changed and positioned, today largely by a scarce, retiring workforce with tongs and manipulators in extreme heat, noise and force. Manual handling caps takt, injures people, mishandles hot parts and cannot meet the rate the aerospace, energy and EV-driveline ramp demands.
Forgebot drives robotic and manipulator handling of hot billets, forgings, dies and trim or inspection parts, perceiving part pose, temperature and grip, planning safe paths around presses and furnaces, and coordinating takt with Heatrix, Forgeon and Flawnex so parts arrive at the right station at the right temperature and time. Skills are trained and validated in Isaac Sim before they deploy to the plant edge, and run under human-approved graduated autonomy with hard safety interlocks.
Automotive and driveline, aerospace and defense, energy and industrial forging makers automating billet, forging and die handling and die changes; forging-automation, robot and manipulator OEMs and integrators.
Named in the product brief as the design
Planned
Training manipulation policies and perception in Isaac Sim, generating sim-to-real data, and running real-time on-robot perception and control are all GPU-bound. CPU-only cannot meet simulation throughput or on-robot control latency for safe hot-part handling.
One robotic pick-and-place handoff, furnace to press or press to trim, with thermal-aware perception validated in Isaac Sim; graduate to multi-cell takt coordination and robotic die change. First metric: transfer time, temperature loss in transfer or manual handling removed on a named cell.
Tier. Cell tier, $15,000 per handling cell or robot per month, expanding to Plant and Enterprise.
Open questions
Hit soundness, dimension and takt before the stroke: the as-forged digital twin.
Forging decisions are made blind and late. Whether a part fills, aligns grain flow, avoids laps and cracks, hits dimension after cool-down and meets takt depends on a coupled chain, billet heat, press stroke, die load, trim and heat-treat, that today is tuned by trial forgings, offline simulation runs and craft. Process development burns heats, dies and weeks; new part introduction and alloy changes are slow and expensive; and forge managers cannot see yield, energy and takt trade-offs before committing the line.
Forgetwin simulates billet heating, die thermal state, press stroke, fill, stress, die load, trim and heat-treat, as-forged against as-designed, to hit target soundness, dimension and properties before the stroke, then sequences and balances the moving line, die allocation, heating batches, press and heat-treat queues and robot takt, for maximum yield at minimum scrap, die wear and energy. It auto-optimises the setpoints that Heatrix, Forgeon and Diesmith execute and that Flawnex verifies, learning from every real outcome.
Forging engineering and operations teams introducing new parts, changing alloys or ramping capacity across automotive and driveline, aerospace and defense, energy and industrial forging; forging-simulation and MES vendors.
Named in the product brief as the design
Planned
Forging FEM, metal-forming, thermal and die-stress, Omniverse twin simulation, physics-informed surrogate training and large-scale sequencing optimisation are classic accelerated-computing workloads. CPU-only cannot run the twin fast enough to inform decisions before the stroke.
As-forged simulation and pre-stroke fill and defect prediction for one part family using existing forging-FEM data; add yield and takt sequencing; graduate to closed-loop setpoint auto-optimisation. First metric: new-part introduction time or ramp scrap on a named part family.
Tier. Sold inside Plant ($90,000 per month) and Enterprise as the twin and optimisation layer.
Open questions
Products are licensed per cell on the Cell tier and bundled on Plant and Enterprise.
| Capability | Cell | Plant | Enterprise |
|---|---|---|---|
| Cells covered | 1 | All on site | All in group |
| Agent modules | 1 of 6 products | All agents | All agents + custom models |
| As-forged twin (Forgetwin) | No | Included | Custom part-family twins |
| Plant-edge node | Included | Per cell | Fleet managed |
| Guarded controller writes | Scoped tags | All scoped tags | Group policy |
| Forging genealogy | Included | Included | Included + qualification packets |
| Autonomy levels | L1 to L2 | L1 to L3 | L1 to L4 when qualified |
| SSO and RBAC | No | Planned | Planned, with approval chains |
| On-prem deployment | No | Optional | Included |
| Support | Business hours | 24/7 | 24/7 + named engineer |
| Indicative price | $15,000/mo | $90,000/mo | $650k to $8.5M ACV |
Every Forgiron run is an ordered, inspectable sequence. This is run_7c2e41 on Cell F-2 · 2,500 t closed-die press · 42CrMo4 connecting rod. Scenario completes in 10 steps with one approval gate. Surface temperature held at 1,205 °C with a 14 °C core-to-surface spread, fill predicted and confirmed complete, one lap precursor at the flash land cleared by a blow-energy trim, lot L-80412 released with full genealogy.
Plant Orchestrator · Pulled the part spec for the 42CrMo4 connecting rod from MES: forging window 1,180 to 1,230 °C, two-blow closed-die schedule, flash land 2.4 mm, temper to 28 to 32 HRC, 1,200 pieces. Locked the qualified process envelope.
Forge Twin · Simulated 24 candidate heat-and-stroke recipes on the as-forged twin for soak time, blow energy and spray timing, ranked on predicted fill, lap risk at the flash land, die load and kWh per piece.
Heat-and-Billet · Induction power on IH-2 stepped down 4% in the final soak stage as the pyrometer read 1,205 °C at the surface. Core-to-surface spread estimated at 14 °C against a 20 °C limit, no overheat, scale growth inside the window.
Die-and-Lube · Lower-die spray dose raised 6% and cooling extended 0.3 s as die surface temperature climbed through 240 °C. Heat-check growth on cavity 2 flagged at 0.31 of the die-change threshold.
Plant Orchestrator · Recipe #9 moves second-blow energy 3% below the standing schedule. Press-schedule changes stay supervised on this part family, so the step was held for the shift forging engineer and approved at 05:47:31.
Forge-and-Press · Two-blow schedule executed with the approved energy. High-speed vision and force curves confirmed complete fill on the first ten pieces; predicted lap risk at the flash land fell from 0.27 to 0.08.
Defect-and-Inspect · Eight cameras, two thermal cameras and the ultrasonic station fused at the edge. One lap precursor at the flash land on piece 212 flagged and cleared by a 1.5% blow-energy trim inside the approved envelope; the piece was routed to MPI and passed.
Robot-and-Handling · Press-to-trim transfer planned and executed by the handling robot at 11 s per piece, temperature loss in transfer 38 °C against a 60 °C limit, interlock zones clear throughout.
Trim-and-Heat · Temper furnace HT-1 zone 2 trimmed 5 °C against the predicted hardness from the recorded heat and stroke history. First hardness checks read 30 HRC, inside 28 to 32.
Quality-and-Conformance · Lot L-80412 released: genealogy written (billet lot, heat curve, stroke record, die cycles, inspection map, approvals, temper record) and dispositioned against IATF 16949 criteria with cited evidence for the quality engineer's sign-off.
A stroke lasts a fraction of a second and a billet cools while it waits. Perception has to be local, deterministic and fast, so Forgiron runs GPU inference at the plant edge and keeps training, simulation and optimisation in the cloud or on-prem.
8 to 24 cameras per line plus thermal, ultrasonic, MPI, press and ram streams fused at the cell. Sub-100 ms visual alerts and sub-500 ms recommendations are the design targets. Jetson Thor is planned for higher sensor density.
Defect, ultrasonic and MPI fusion, die-wear, heating, press-fill, property and disposition models behind one orchestrator at the plant edge, routed by latency, tenant, sensitivity and cost, with canary rollout by cell.
Multimodal defect, die-wear, heating, press-fill, property and quality-disposition models trained on thermal images, high-speed vision, ultrasonic and MPI signals, press curves and engineer corrections. Dedicated training capacity is planned.
Billet heating, die thermal state, press stroke, fill, stress, die load and robot handling simulated before the stroke. GPU forging FEM is the workload; physics-informed surrogates are planned so recipes rank at production cadence.
Rare lap, fold, underfill, crack and die-wear scenarios generated and validated against real inspection distributions before promotion. Planned.
Die allocation, press sequencing, heating batches, heat-treat queues, robot and takt coordination and inspection routing.
Hot billet, forging, die and trim handling skills trained and validated in simulation before they run next to a press.
Daily features from thermal, video, ultrasonic, MPI, press and genealogy data for retraining, drift detection and yield, scrap, die-life and energy benchmarks.
Land on one workflow with a measured baseline. Expand by cell, module and autonomy level. Outcome components are available on yield, scrap, die life and energy. Annual prepay carries a 15 to 20% discount.
Heating or forging control, die management, or inspection AI for one cell. For forges validating ROI on the wedge.
The whole forge: heat and billet, forge and press, die and lube, trim and heat-treat, defect, quality and property, yield and takt, robotic handling and the twin.
Multi-site forging groups standardising on Forgiron, with custom alloy, die and qualification models.
The questions plant directors, forging engineers and quality engineers ask in the first meeting.
Yes, but only within an explicit tag allow-list with per-tag rate and magnitude limits, inside the qualified process envelope, and only at the autonomy level your plant has set. Design partners start in shadow mode, where Forgiron writes nothing. Press-schedule changes stay supervised by a named forging engineer until the part family is qualified.
Control returns to the furnace, press or die controller's last known-good state. Forgiron is a supervisory layer on top of the PLCs and interlocks you already run, never a replacement for them, so an outage degrades the cell to its current way of running, not to a stop.
Advisory recommendations start from the physics-based twin and the part spec in the first weeks. Defect, die-wear and heating prediction improve as cell history and labelled outcomes accumulate per part family. The first training sets are sized at 2 million labelled frames and signals, 20,000 heat and press cycles and 10,000 die-life events per design-partner cohort, all design targets.
Only if you choose cloud training. Die designs, part drawings, recipes, process models and quality records are isolated per tenant and never used to train another customer's models. A fully on-prem deployment at the plant edge is available for sensitive producers, and federated learning across sites is planned and waits on data-rights agreements.
You are, the same as with any control strategy, which is why every write is policy-checked, bounded, logged and reversible, and why anything above your risk threshold waits for a named approver. The audit log records the request, the evidence, the envelope applied and the human decision.
A paid pilot of 90 to 180 days on one cell, scoped to a single wedge with a success metric written into the order form: defect-detection accuracy, scrap reduction, die-change reduction, energy reduction or first-pass-quality improvement. The first 30 to 60 days baseline scrap, die life, energy, inspection escapes and takt in shadow mode.
Start with one heating, press, die or inspection cell and one measurable baseline. A paid pilot on billet heating and press fill, die life or ultrasonic and MPI defect detection shows the number before you commit further.
↑↓ navigate↵ openesc close
