Per-tenant everything
Data, models, retrieval indexes and memory are scoped per tenant and per entity. On-prem deployment is available for sensitive producers.
FAQ
Answers beforethe first call.
The questions plant directors, process engineers, quality managers, IT and OT reviewers and procurement teams ask us — answered directly, including the parts that are still aspirational.
The basics
What Extruonactually is.
It closes the loop on extrusion. Extruon senses the billet, profile, surface and temper with fused vision, thermal and laser gauging; plans coupled moves across billet heating, press, die, quench, stretch, cut and aging; acts inside approved envelopes at the plant edge; and logs every action to an assurance-grade audit trail that feeds a training loop.
A press PLC holds setpoints but is blind to profile quality and recovery drift. A gauge measures but does not act on the press. A flow simulator runs offline before the die exists. Extruon unifies perception, decision and action into one plant-edge platform that acts on the process and compounds the resulting data.
No. There is a console, but the product is a system of action. If all you want is another visualisation layer, we are the wrong vendor and will say so on the first call.
Aluminium extrusion first, where the physics, the tolerances and the economics are best understood. Polymer and plastic profile, tube and co-extrusion are on the roadmap. [ASPIRATIONAL]
We mark aspirational claims throughout the site. In short: the architecture, connectors, agent loop, review console and audit log are the build focus; multi-site fleet learning, full robotic-handling autonomy and published audited case studies are ahead of us.
Autonomy & safety
How softwareearns a press.
Only when you enable it. Deployments start in shadow mode with no write path at all, progress to assist mode where engineers approve each recommendation, and only reach bounded autonomy on the presses and setpoints you approve, inside envelopes your engineers define, with rollback on any alarm.
Four independent mechanisms: action envelopes with min, max and max-step per setpoint; human approval checkpoints on high-impact and temper-critical decisions; twin validation and golden-dataset gates in CI before promotion; and alarm interlocks wired into your existing fail-safe stop.
The affected agent degrades to advisory rather than acting on partial perception. The degradation appears in the console and the audit trail, and the runbook defines the checks before resuming.
Your engineers define the envelopes and approve the autonomy level; Extruon logs every decision with its evidence, model version and approver. Accountability is explicit in the audit record rather than argued after the fact.
Yes. Autonomy level is reversible per press at any time, and any alarm reverts the last move and hands control back automatically.
The short answer
It is allabout one push.
Whatever the question, the answer usually reduces to: what did the system see on this push, what did it decide, was that inside the envelope, and what happened to the metal?
Perception, decision, envelope, outcome — recorded together, every time.
Security & data
What leavesyour boundary.
Die geometry, customer drawings, press recipes, order book, plant performance and quality records never leave your tenant boundary. Fleet learning is opt-in per cohort and shares only anonymised defect signatures and aggregate process priors.
Yes. Full on-premise deployment is available with the same feature set. Model updates then arrive through your own controlled channel instead of an OTA link.
SOC 2 Type I is in progress, with Type II targeted as enterprise deployments mature. [ASPIRATIONAL] Current controls, policies and the security pack are available under NDA.
Access is role-based and scoped per tenant, with permission-aware retrieval filtering. Support access is time-bounded, logged and requires your authorisation.
The perception snapshot, model and version, the proposal and its rationale, the envelope check, the approver and the measured outcome — for every action, exportable as Parquet to your own storage.
Integration
What it takesto connect a plant.
Ram speed, pressure and position plus container and die temperatures from the press, and zone setpoints from the billet-heating furnace, over OPC UA or Modbus TCP. Quench, stretcher, saw, aging oven and MES connections extend the loop.
We deploy press-exit sensing — line-scan and RGB cameras, thermal imaging and laser gauging — as part of the wedge. Existing gauges are integrated where they exist.
Connector work is typically days for a press with clean OPC UA access. Sensor installation at the press exit is usually the long pole, followed by a shadow-mode period of several weeks before any recommendations are acted on.
Common surfaces are supported over REST, OPC UA and MQTT. Long-tail integrations beyond design-partner stacks are explicitly out of scope in v1. [PLACEHOLDER — current connector list on request.]
Sensing installation is scheduled into planned downtime. Shadow mode has no write path, so the loop runs against live production without touching it.
Commercial
Pricing, pilotsand expansion.
Press is $12,000 per press line per month for press and heating control or profile and defect sensing on one press. Plant is $78,000 per month for whole-plant autonomy including the twin. Enterprise is custom, landing $600k to $7M ACV. Annual prepay carries roughly a 15 to 20 percent discount.
No. Pilots are paid with a written success metric. A paid pilot produces an owner, a baseline and a decision date; a free one produces polite interest.
You have spent a quarter, not a capital programme, and you keep the baseline work. We would rather that than a long deployment nobody can evaluate.
One press on one module, then the same module across presses, then adjacent modules on the proven press, then the second site. Expansion is the growth engine — target NRR 135%.
Edge compute and sensing hardware are quoted separately at cost, since requirements vary by press, sensor count and existing gauging. [PLACEHOLDER — hardware price list on request.]
Measured on the press
The numbersbehind the answers.
Every Extruon engagement starts with a baseline and ends with an audited delta. These are the target bands we underwrite in a paid pilot.
+4.2 pts
Press recovery uplift, saleable vs charged metal
−38%
Surface-defect and dimensional rejects
−61%
Die-trial pushes before a die is signed off
−12%
kWh per tonne across heating, quench and aging
Target outcome bands modelled from design-partner baselines. [ASPIRATIONAL — to be replaced with audited pilot results.]
Assurance-grade by default
Trust is the gateto autonomy.
Extruon never asks for control it has not earned. Shadow mode proves accuracy against your operators and gauges; advisory mode proves ROI; bounded autonomy only follows once the gates are met.
Your die geometry stays yours
Die designs, customer drawings and recipes never leave your boundary. Fleet learning shares defect signatures and process priors, never geometry.
Immutable action log
Every perception, recommendation, approval and setpoint write is recorded with its evidence, ready for quality and customer audits.
SSO, RBAC, encryption
SSO and role-based access, encryption in transit and at rest, SOC 2 programme in progress, and fail-safe interlocks on every write path.
Seven agents, one orchestrator
The plant runs as a crewof specialised agents.
Each agent owns a physical stage of the line, senses it directly, and acts inside an approved envelope. A plant orchestrator arbitrates between them so nobody optimises recovery at the expense of temper.
Heat-and-press agent
Controls billet temperature taper, ram speed, breakthrough pressure, container and die temperature, and exit temperature to keep metal flow balanced and the profile on dimension.
Profile-and-defect agent
Senses and predicts cross-section dimensions, wall thickness, straightness and surface defects — die lines, pickup, scoring, blisters — from fused vision and laser gauge.
Die-and-flow agent
Optimises die temperature, correction and flow balance to cut die trials and the twist and bow that send profiles back to the stretcher.
Quench-and-stretch agent
Controls press-quench cooling rate and stretch-straightening so temper and flatness land inside spec on the first pass.
Cut-and-age agent
Controls cut-to-length, batching and aging ovens to reach T5 and T6 mechanical properties without over-soaking the oven.
Robot-and-handling agent
Drives robotic pulling, stacking, racking and packing of long, hot, delicate profiles — the handling nobody wants to staff at 3am.
Yield-and-throughput agent
Optimises press recovery, scrap, butt and offcut loss, and energy per tonne across the whole line rather than one station at a time.
Extrusion-knowledge agent
Answers metallurgy and die-design questions with citations into your die books, profile drawings, press recipes and alloy specs.
Press-and-die digital twin
Simulates billet heating, metal flow through the die, quench and profile properties — hitting the target profile and yield before the push, not after the scrap.
Extruders asking these questions
NORDHAUS ALUVantera ProfilesSolFrame SystemsKestrel ExtrusionsRhein LeichtbauAurora Billet Co.
Design-partner names are illustrative pending public reference approval. [PLACEHOLDER]
FAQ
Still havea harder question?
Ask an engineer directly. The awkward questions about safety, IP and evidence are the ones we most want to answer.
Land on one press. Expand press by press, module by module, site by site.