The orchestrator plans, the agents act
A run starts as a goal, becomes a plan, and executes as a graph of typed steps. The active path is always visible, always logged, always reversible.
Draweon is an agent operating system for wire and cable plants. Specialist agents control drawing and annealing, stranding and cabling, and insulation and jacket extrusion, watched by fused laser, eddy-current, spark and vision inspection, rehearsed on an as-run line twin, and bounded by a policy engine you control.
Illustrative scenario: EX-3 · 2.5 mm² class 2 PVC building wire, changeover, no spark faults
Design partners wanted · 3 to 5 plants · lines we are looking for
Pre-launch. No customers, pilots or logos are claimed. The first cohort gets a paid pilot on one line against a baseline agreed in writing.
Every Draweon run is an ordered, inspectable sequence. This is run_3d7a1e on EX-3 · 2.5 mm² class 2 PVC building wire. Scenario completes in 10 steps with one approval gate. Wall settles at 0.80 mm nominal, minimum 0.66 mm, concentricity 0.93, no spark faults, reel R-30417 released with full genealogy.
Plant Orchestrator · Pulled the construction spec for 2.5 mm² class 2 PVC building wire from MES: 7 × 0.67 mm conductor, 0.80 mm nominal wall, 0.62 mm minimum spot wall, 6 kV spark test, 48 km order. Locked the target envelope.
Line Twin · Simulated 36 candidate recipes on the as-run line twin for draw schedule, lay length and extrusion setpoints, ranked on wall variance, concentricity, predicted spark risk and polymer give-away.
Draw-and-Anneal · Capstan tension on RB1 stepped down 6% as the die-wear estimate crossed its threshold. Diameter held at 0.670 mm ± 0.004 mm and the break-risk score fell from 0.31 to 0.09.
Draw-and-Anneal · Annealer current trimmed against the live conductivity reading to hold the softness target without over-anneal through the speed change.
Strand-and-Cable · Lay length set to 28 mm on bunching line ST2 with payoff tension balanced across 7 spools. Roundness index held above 0.97 into the extrusion payoff.
Insulate-and-Jacket · Screw speed and zone 3 temperature re-cut against melt pressure and material-lot moisture. Wall settled at 0.80 mm nominal, minimum 0.66 mm, concentricity 0.93.
Defect-and-Inspect · 6 cameras, the laser gauge, the eddy-current head and the spark tester fused at the edge. One neckdown precursor at 12,480 m flagged and cleared by a screw-speed trim before it reached the reel.
Plant Orchestrator · Line-speed ramp 1,100 to 1,250 m/min exceeded the site autonomy threshold. Held for the shift process engineer and approved at 06:41:52.
Robot-and-Handling · Reel change on take-up TU3 planned and executed by the handling robot. Payoff aligned and tension re-established inside the changeover window.
Quality-and-Conformance · Reel R-30417 released: genealogy written (conductor lot, draw and anneal record, lay, wall and concentricity trace, defect map, approvals) and checked against the construction spec.
Every tool invocation, argument and result is written to an append-only, human-readable log, and every reasoning step is expandable. Nothing about a run is hidden from the plant.
Plan, thought, action, observation, exactly as the orchestrator would record it.
Change line EX-3 over to 2.5 mm² class 2 PVC building wire, hold 0.80 mm nominal wall and concentricity, and finish the run with zero spark faults.
Simulate the changeover on the twin first. Lead with drawing and annealing so the conductor is right before stranding, then set lay, then re-cut extrusion against the live wall reading. Fuse inspection continuously. Hold the line-speed ramp for a human.
The standing recipe raises line speed and screw speed together. On the twin that produces a wall dip at zone 3 because the melt pressure lags the speed step. Sequencing the zone 3 temperature move ahead of the speed ramp removes it.
twin.simulate(candidates=36) returns recipe #22: zone 3 up 4 °C before the ramp, screw speed trimmed to the live wall reading, lay 28 mm.
Recipe #22 predicts wall variance 0.018 mm, spark risk 0.04 and give-away 2.1%. Three candidates scored lower on give-away but pushed predicted minimum wall under the spec floor and were discarded.
Execute recipe #22 under autonomy level L3: eight setpoint writes permitted, the line-speed ramp routed to the shift process engineer.
Neckdown precursor at 12,480 m at 06:39:04. A 1.5% screw-speed trim cleared it in 16 s with no reel quarantined.
Scenario complete. Wall 0.80 mm nominal, minimum 0.66 mm, concentricity 0.93, zero spark faults. Reel R-30417 released with full genealogy.
Nothing about a Draweon run is implicit. The goal is declared, the plan is simulated on the twin, each step is typed and bounded, and the outcome is written to the reel record.
A run starts as a goal, becomes a plan, and executes as a graph of typed steps. The active path is always visible, always logged, always reversible.
L1 advisory, L2 supervised writes, L3 bounded autonomy, L4 unattended within an envelope. Each plant sets the level per agent, per tag, per shift, and every write above the threshold waits for a named approver. The four figures above are design targets, not measured results.
Each agent owns a section of the line, a bounded tool set and a metric the plant already measures. They negotiate for shared actuators through the orchestrator, never directly. Five of them ship as named products.
Rod breakdown, wire drawing, annealing
Perceives rod, wire, tension, die force, temperature and annealer conductivity in real time, predicts break risk and die wear before they happen, and closes the loop on die and capstan sequencing, tension, line speed and annealing setpoints to hold diameter and conductivity.
Stranding, cabling, lay and geometry
Senses payoff tension, take-up, rotational speed and geometry, predicts lay and roundness drift and birdcaging before they form, and closes the loop on strander and take-up speeds and tensions to hold target lay length and geometry.
Insulation and jacket extrusion
Fuses gauge, laser and wall-scan signals with melt and cooling telemetry, predicts eccentricity and wall drift before it lands on the reel, and closes the loop on screw speed, temperature zones, line speed and cooling to hold concentric wall and dielectric quality.
Fused inspection and defect prediction
Fuses vision, laser, eddy-current and spark-test signals into one real-time model of cable health. Detects and predicts diameter drift, lumps and neckdowns, insulation faults, sparks and pinholes, localises them to reel and length, and routes bounded corrections upstream.
Conductor, insulation and dimensional quality
Holds conductor, insulation and dimensional quality against the construction spec across the whole line, reconciling what the control agents want with what the spec allows.
Yield, scrap, length give-away, energy, line balancing
Prices every candidate move in scrap, give-away, kilowatt-hours and minutes, sequences changeovers and reel changes, and balances line speed against downstream capacity.
Spool, reel, payoff and take-up handling
Drives spool, reel and payoff and take-up handling so the line never waits on a reel change, with motion planned and checked in simulation before it runs near moving conductor.
Right-first-time, non-conformance, genealogy, traceability
Builds the reel and length genealogy record, gates release against the construction spec and the standards the cable is qualified to, and assembles evidence packets for customer and regulatory review.
Specs, standards, procedures, operator Q&A
Answers process and standards questions from draw schedules, construction and insulation specs, cable standards and quality procedures, with every answer cited to its source and scoped to the tenant.
As-run whole-line simulation and recipe optimisation
Simulates rod breakdown, drawing, annealing, stranding, extrusion, cooling and reel handling with GPU physics, compares as-run against as-designed, and auto-optimises draw schedule, lay and extrusion recipes so the line hits target quality, dimension and takt before the first metre runs.
Planning, arbitration, human approval
Plans the run, arbitrates between agents competing for the same actuator, enforces the autonomy level and routes anything above the risk threshold to a named approver.
Draweon is not a dashboard. Each agent closes a loop a cable plant already measures in scrap, metres, kilowatt-hours and minutes. The tables show what each loop watches and what it is allowed to move.
Draw tension, die force and temperature are fused at the edge into a rolling break-risk estimate. When risk crosses the threshold the agent has time to trim tension or step the die sequence instead of restringing a line.
| Signal | Watches | Moves |
|---|---|---|
| Draw tension | Watched per capstan | Trimmed inside per-tag limits |
| Die wear | Estimated from force and temperature | Die-change recommendation |
| Break risk | Scored continuously | Line-speed hold or tension trim |
Walls run thick because thin is dangerous. Extrio fuses gauge, laser and wall-scan with melt and cooling telemetry and holds the wall at nominal, not at nominal plus a margin, while keeping minimum spot wall above the floor.
| Signal | Watches | Moves |
|---|---|---|
| Wall thickness | Gauge, laser and wall scan | Screw speed and line speed |
| Concentricity | Wall scan per quadrant | Die centring recommendation and cooling |
| Melt pressure | Head and zone telemetry | Zone temperatures |
A spark tester tells you about a fault after it is on the reel. Eddix fuses vision, laser, eddy-current and spark signals to score insulation health ahead of the tester and route a bounded correction upstream.
| Signal | Watches | Moves |
|---|---|---|
| Spark events | Tester plus upstream precursors | Recommendation to Extrio, reel quarantine |
| Neckdowns and lumps | Laser and cameras | Screw-speed trim via Extrio |
| Eccentricity | Laser and wall scan | Die-centring and cooling recommendation |
Every changeover and new construction is simulated on the as-run line twin before a setpoint moves. Candidates that save polymer but push minimum wall under the floor or raise spark risk are discarded automatically.
| Signal | Watches | Moves |
|---|---|---|
| Draw schedule | Die train and tension plan | Proven setpoints to Drawix |
| Lay length | Stranding geometry | Proven setpoints to Cablyn |
| Extrusion recipe | Screw, zones, cooling, speed | Proven setpoints to Extrio |
Two agents will want the same actuator. The orchestrator arbitrates on the run goal, not on who asked first, and the handoff is logged like any other step.
Yield-and-Takt wants to trim screw speed to reduce polymer give-away on the thick side of the wall.
Defect-and-Inspect wants to raise screw speed to clear a neckdown precursor at 12,480 m.
Neckdown risk 0.58 outranks 0.3% give-away under the run goal "zero spark faults". Defect-and-Inspect wins the actuator for 120 s.
Actuator returned. Yield-and-Takt recovers give-away through the cooling trough instead. Wall holds at 0.80 mm nominal.
Illustrative scenario.
A plant does not go from manual to unattended in one step. Draweon makes the level explicit, auditable and reversible at any time. The first release runs shadow mode, then assist mode, then graduated autonomy.
| Level | What the agent does | What the human does | Where it sits in a pilot |
|---|---|---|---|
| L1 · Advisory | Recommends setpoints and explains why, with citations | Enters every change manually | Shadow mode, weeks 1 to 3 |
| L2 · Supervised | Proposes a write; it executes on approval | Approves each write in the review console | Assist mode, weeks 4 to 8 |
| L3 · Bounded | Writes inside tag, rate and magnitude limits | Approves ramps, die changes and reel release | Graduated autonomy, from week 9 if the plant is satisfied |
| L4 · Unattended | Runs the envelope without prompting | Sets the envelope; reviews the shift record | Planned; after qualification on the construction |
These are the figures in our compute plan. None is a measured result. Every pilot measures against the plant's own baseline, reproduced from the plant's reel genealogy, not from our telemetry.
Power, building wire, automotive and EV harness, data and fibre, magnet and specialty. Most plants pick one line and one wedge, prove the number, then expand.
Large conductors, long runs and dielectric requirements that leave no margin. Drawing, stranding and insulation control with fused inspection on every reel.
Thin walls at high line speed, where give-away and spark faults decide margin. Extrio and Eddix on the insulation line first.
Small conductors, thin walls and USCAR qualification. Concentricity and spark prediction before the reel, genealogy for every lot.
Lay and geometry decide electrical performance. Cablyn holds lay and roundness and hands a round core to extrusion.
Fine diameters, exact conductivity and softness. Drawix on break prediction and annealing control, energy per reel measured.
Conductor moves at over a thousand metres a minute. Perception has to be local, deterministic and fast, so Draweon runs GPU inference at the plant edge and keeps training, simulation and optimisation in the cloud or on-prem.
4 to 12 cameras per line plus laser, eddy-current and spark streams fused at the line. Sub-50 ms alerting and sub-250 ms bounded control are the design targets. Jetson Thor is planned for higher sensor density.
Vision, time-series, anomaly and quality models behind one orchestrator at the plant edge, with canary rollout by line and fallback to the previous validated version.
Defect classifiers, sensor-fusion models, time-series prognostics and cable-process reasoning models trained on reel telemetry and engineer corrections. Dedicated training capacity is planned.
As-run drawing, stranding, extrusion, thermal and tension simulation. Physics-informed surrogates are planned so recipes can be optimised at production cadence.
Rare spark, pinhole, eccentricity, neckdown and break variants generated and validated against real gauge, spark and vision distributions before promotion. Planned.
Line sequencing, reel changeovers, energy allocation, takt and constrained setpoint proposals.
Spool, reel and payoff and take-up motion planned and validated in simulation before it runs next to moving conductor.
The Draweon SDK is typed Python. Tools are declared with schemas and limits and the policy engine enforces them at call time, not in a review meeting. This is a design preview; the API will change before release.
# Design preview. Bound the extrusion agent to line EX-3.
from draweon import Agent, Tool, Limit, Autonomy
screw = Tool(
name="plc.write_setpoint",
tags=["EX3.SCREW_RPM", "EX3.ZONE3_T"],
limits=[Limit(max_step="2%", per="30s")],
)
extrude = Agent(
id="agent.insulate_jacket",
goal="wall 0.80 mm nominal, min 0.62, concentricity >= 0.90",
tools=[screw, Tool("gauge.read_wall", read_only=True)],
# bounded writes; humans still gate ramps
autonomy=Autonomy.L3,
# simulate on the twin before every write
verify="twin",
)
run = extrude.start(line="EX-3", construction="BW-2.5-C2-PVC")
for step in run.stream():
print(step.name, step.status, step.duration)
The same run engine, the same policy checks, the same audit trail, from the terminal, the review console or the SDK. Design preview.
$ draweon run "changeover EX-3 to BW-2.5-C2-PVC" --autonomy L3
→ plan composed 10 steps · 1 approval gate
→ twin.simulate 36 candidates · best #22 · spark risk 0.04
→ policy.evaluate 8 writes permitted · 1 held for human
→ executing draw.tension ..... ok 1m48s
→ executing anneal.current ... ok 54s
→ executing strand.lay ....... ok 1m12s
→ executing extrude.wall ..... ok 2m36s
! approval required line_speed 1100 → 1250 m/min
→ approved shift process engineer · 06:41:52
→ scenario complete spark faults 0 · reel R-30417 released
$ draweon runs show run_3d7a1e --format genealogy
Land on one workflow with a measured baseline. Expand by line, module and autonomy level. Outcome components are available on yield, scrap, length and energy. Annual prepay carries a 15 to 20% discount.
Drawing or extrusion control, or inspection, for a single line. For cable makers validating ROI on the wedge.
The whole loop: draw and anneal, strand and cable, insulate and jacket, defect, quality and dimension, yield and takt, robotic handling and the twin.
Multi-site wire and cable groups standardising on Draweon, with custom conductor, insulation and tolerance models.
We have no customer quotes to show yet and will not invent any. These are the three roles in a cable plant we build for, and the problems they bring to the first meeting.
What this person is dealing with today
What this person is dealing with today
What this person is dealing with today
Draweon reads and writes through the systems already on the floor. No rip and replace, no parallel historian, no new HMI to learn. Vendor names are examples of systems found on cable lines; connector availability is confirmed per site.
PLC and SCADA on drawing, stranding and extrusion lines (Siemens, Rockwell and OEM controllers)
Setpoint reads and guarded writes
Laser gauges and X-ray wall scanners (Sikora, Zumbach, Beta LaserMike and NDC class systems)
Wall, diameter, eccentricity traces
Eddy-current heads and spark testers
Fault events with length position
Line-scan and area cameras on the line
Frames for lumps, neckdowns, surface faults
Rod breakdown, intermediate and fine drawing, annealers (Niehoff and Properzi class lines)
Tension, die force, annealer current, conductivity
Bunchers, planetary, rigid and tubular stranders (Gauder, Pourtier, Nextrom class lines)
Payoff tension, take-up, lay, geometry
Insulation and jacket lines (Maillefer, Rosendahl Nextrom, Troester, Davis-Standard class lines)
Screw, zones, melt pressure, cooling
Cable MES, quality systems and in-house records
Orders, constructions, reel genealogy
Draweon writes to production equipment. Every capability is scoped, every write is policy-checked, and every action is written to an append-only audit log the plant owns.
| Item | Scope | Status |
|---|---|---|
| SOC 2 Type I | Cloud control plane | QUEUED Planned, first 6 months |
| SOC 2 Type II | Cloud control plane | QUEUED Planned, 6 to 12 months |
| SSO and RBAC | Console and approvals | QUEUED Planned, 6 to 12 months |
| Standards-cited reasoning | UL, IEC, IPC, USCAR documents | RUNNING In the design, every answer cites its source |
| Per-tenant isolation | Recipes, models, memory | RUNNING In the design from day one |
| On-prem and plant-edge deployment | Sensitive producers | RUNNING Design option |
One policy model, one audit trail, one benchmark across every line in every plant, with the conductor, insulation and tolerance models kept private to each site.
The questions plant directors and process engineers ask in the first meeting.
Yes, but only within an explicit tag allow-list with per-tag rate and magnitude limits, and only at the autonomy level your plant has set. Level 1 is advisory: Draweon recommends and a human enters everything. Design partners start there, in shadow mode, before supervised writes.
Control returns to the line controller's last known-good state. Draweon is a supervisory layer on top of the PLCs and gauges you already run, never a replacement for them, so an outage degrades the line to its current way of running, not to a stop.
Advisory recommendations start from the physics-based twin and the construction spec in the first weeks. Defect prediction and break prediction improve as line history and labelled events accumulate per construction. The first training sets are sized at 5 to 20 million labelled frames and events, a design target.
Only if you choose cloud training. Recipes, process models and defect libraries 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.
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 reasoning, the limits applied and the human decision.
A paid pilot on one line, scoped to a single workflow with a baseline agreed in writing. Weeks 1 to 3 are connection and shadow-mode observation, weeks 4 to 8 are assist mode with engineer approval, and bounded autonomy follows if the plant is satisfied with the recommendations.
Start with one drawing, stranding or extrusion line and one measurable baseline. A paid pilot on wall and concentricity, break prediction or spark prediction shows the number before you commit further.
↑↓ navigate↵ openesc close