Each Glasent product is an agent with a section of the plant, a bounded tool set and a measured outcome. They negotiate for shared actuators through the orchestrator and never write outside their allow-list.
Each product owns a section of the glass plant, a bounded tool set and a measured outcome. They negotiate for shared actuators through the orchestrator, never directly.
Batching, furnace melting, pull rate, melt chemistry
Autonomous batching and furnace melting, held on chemistry every second. Real-time batch, melt-chemistry and homogeneity modelling with closed-loop furnace control.
Watches: Furnace zone temperatures, pull rate, batch weights, redox and fining signals, lab chemistry, Seedscan seed and stone rate
Float-bath and IS-machine forming, gob weight, dimension
Autonomous float-bath and IS-machine forming, dialled to exact dimension. Real-time gob-weight, pull-rate and dimension modelling with closed-loop forming control.
Watches: Gob weight and temperature, ribbon speed, top-roller angle, tin-bath conditions, thickness and flatness scans, Seedscan check risk
Lehr zones, cooling curve, residual stress
Autonomous lehr annealing that relieves stress before it becomes breakage. Real-time lehr-zone and residual-stress modelling with closed-loop cooling control.
Watches: Lehr zone temperatures, belt speed, cooling rate, birefringence and polariscope readings, product thickness from Formeon
Seeds, stones, cords, checks, stress at line speed
See every seed, stone, cord and stress before it escapes the plant. Multi-sensor fusion of vision, optical, thermal and stress or birefringence at the edge.
Watches: Every camera, optical, thermal and stress station on the line, at 30 to 60 frames a second
Pick, transfer, orient and stack hot, fragile ware
Autonomous handling of hot, fragile glass: panes, ware and stacks. Vision-guided robotic pick, transfer, orient and stack for panes, ware and substrates.
Watches: Cell cameras, ware geometry and temperature, Seedscan defect flags per piece, robot and conveyor state
Simulate melt, form and anneal; optimise before the run
Hit target quality, stress and takt before the run, in simulation first. GPU-accelerated furnace-CFD, glass-flow and annealing-stress simulation of the full chain.
Watches: The as-run state of the furnace, forming line and lehr, plus every candidate recipe the orchestrator proposes
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 glass technologist.
100% of writes policy-checked
The six products are the first agents. The architecture names four more scopes that ship inside the Plant tier as the downstream lines are connected.
Coating, tempering and finishing to exact spec. Ships inside the Plant tier as the lines downstream of the lehr are connected.
Thickness, flatness, dimension and grade optimisation across the line, built on Seedscan and Formeon signals.
Yield, cullet, energy and line balancing, priced in tonnes, megawatt-hours and minutes, solved with cuOpt.
Retrieval over batch recipes, forming and annealing specs, glass standards and site procedures, with a citation on every answer.
Every Glasent run is an ordered, inspectable sequence. This is scenario_fl2_01 on FL-2 · float line · clear soda-lime: Thickness change FL-2 · 6 mm to 4 mm clear float, residual stress inside spec, zero escaped seeds. It is a worked scenario that shows the shape of a run, not a measured customer result.
Plant Orchestrator · Pulled the 4 mm clear float spec, optical-grade tolerances and the standing energy window from plant MES; locked the target envelope for the run.
Glastwin · Simulated 36 candidate transition recipes across furnace pull, tin-bath ribbon speed and lehr curve; ranked them on seed risk, residual stress and energy per tonne.
Meltrix · Stepped furnace pull toward the new ribbon mass flow while holding melt temperature and fining; chemistry stayed inside the composition window.
Formeon · Raised ribbon speed and re-angled the top rollers to thin the ribbon toward 4 mm; thickness converged inside the design tolerance.
Anneon · Re-shaped the lehr cooling curve for the thinner, faster ribbon so residual stress stays inside spec at the higher speed.
Seedscan · 8 camera, optical and stress stations streaming; a seed cluster flagged at the ribbon edge and attributed to the pull transient, routed to cullet.
Plant Orchestrator · Re-sequenced cut sizes so transition ribbon routes to cullet recovery and good ribbon to the highest-value open order.
Plant Orchestrator · The second pull step exceeded the site autonomy threshold. Held for the glass technologist on shift; approved and written to the audit log.
Panebot · Re-planned pick and stack paths for the thinner panes; plates flagged by Seedscan diverted to cullet, good plates stacked to rack A3.
Plant Orchestrator · Lot released with full genealogy: batch, melt, forming, lehr curve, defect map, stress map and the technologist's approval.
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.
Formeon wants pull held while the ribbon thins, to protect thickness convergence.
Meltrix wants the second pull step now, to settle fining before the seed rate climbs.
Seed risk outranks a short thickness excursion under the run goal "zero escaped seeds". Meltrix wins the actuator, and because the step is above threshold it goes to the technologist.
Actuator returned; Formeon recovers thickness with roller angle instead. Both requests, the score and the reason are in the run record.
Every tool invocation, argument and result is written to an immutable, 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.
Move FL-2 from 6 mm to 4 mm clear float with residual stress inside spec and zero escaped seeds.
Simulate the transition on the twin first. Lead with furnace pull so the ribbon mass flow is right before the forming change, then thin the ribbon, then re-shape the lehr curve. Watch the ribbon continuously; hold the second pull step for a human.
The standing recipe changes pull and ribbon speed together. On the twin that produces a fining transient that seeds the ribbon edge. Sequencing the pull step ahead of the forming change and holding fining removes most of it.
twin.simulate(candidates=36) returned recipe #19: pull in two steps, forming change after fining settles, lehr curve re-shaped before belt speed rises.
Recipe #19 scored lowest on seed risk with stress inside spec. Two candidates scored lower on energy but pushed residual stress over the site limit and were discarded.
Execute recipe #19 under autonomy level L3: eleven setpoint writes permitted, the second pull step routed to the glass technologist.
Seed cluster at the ribbon edge at 02:28:41, attributed to the pull transient. cuOpt routed that ribbon to cullet recovery; no flagged plate reached a customer stack.
Scenario run complete. Thickness at 4.0 mm, residual stress inside spec, one approval gate, full genealogy written to the lot record.
Agents do not get a shell. They get a registry of typed tools, each with a direction, units, limits and an owner, enforced at call time.
| Tool | Owner | Direction | Limits |
|---|---|---|---|
| scada.read_furnace | Meltrix | read | Read-only |
| lab.read_chemistry | Meltrix | read | Read-only |
| scada.write_setpoint | Meltrix | write | Allow-listed tags, rate and magnitude limits |
| is.read_gob | Formeon | read | Read-only |
| dim.read_thickness | Formeon | read | Read-only |
| float.write_ribbon | Formeon | write | Allow-listed tags, rate and magnitude limits |
| lehr.read_zones | Anneon | read | Read-only |
| optic.read_birefringence | Anneon | read | Read-only |
| lehr.write_curve | Anneon | write | Allow-listed tags, rate and magnitude limits |
| vision.stream | Seedscan | read | Read-only |
| optic.stress_map | Seedscan | read | Read-only |
| alarm.raise | Seedscan | write | Allow-listed tags, rate and magnitude limits |
| isaac.plan_path | Panebot | write | Allow-listed tags, rate and magnitude limits |
| robot.grip | Panebot | write | Allow-listed tags, rate and magnitude limits |
| wms.route | Panebot | write | Allow-listed tags, rate and magnitude limits |
| twin.simulate | Glastwin | read | Read-only |
| cuopt.sequence | Glastwin | read | Read-only |
| policy.replay | Glastwin | read | Read-only |
| plan.compose | Orchestrator | control | Policy-checked; approvals routed to a named person |
| policy.evaluate | Orchestrator | control | Policy-checked; approvals routed to a named person |
| approval.request | Orchestrator | control | Policy-checked; approvals routed to a named person |
An agent is narrow, auditable and action-bound. Each one has a playbook, allowed tools, hard stop conditions, approval thresholds and replayable logs.
The model router puts the best or cheapest model on each step: fine-tuned open models on high-volume perception and control, frontier models on glass-technology reasoning and question answering. Every output carries a citation to the recipe, standard or procedure it rests on, and continuous evaluation against golden datasets gates every model and prompt change in CI.
A plant does not go from manual to unattended in one step. Glasent makes the level explicit, auditable and reversible at any time, and the first release plan is shadow, then assist, then graduated autonomy.
| Level | What the agent does | What the person does | When |
|---|---|---|---|
| L1 · Shadow and advisory | Observes, predicts and recommends setpoints with its reasoning | Enters every change manually; a baseline is measured | Pilot weeks 1 to 3 |
| L2 · Assist | Proposes a write; it executes on approval | Approves each write in the review console | Pilot weeks 4 to 8 |
| L3 · Bounded | Writes inside tag, rate and magnitude limits on low-risk loops | Approves pull steps, grade releases and anything above threshold | Pilot week 9 onward |
| L4 · Unattended | Runs the approved envelope without prompting | Sets the envelope; reviews the shift record | Planned, after graduated autonomy proves out |
Which agent owns which part of the plant, and what it hands to the next one.
Batching, melt temperature, pull and chemistry. Hands a homogeneous melt to Formeon.
Float-bath or IS-machine forming, gob weight and dimension. Hands dimension-correct ware to Anneon.
Cooling curve and residual stress. Hands stress-relieved ware to Panebot.
Seeds, stones, cords, checks, stress and dimension drift. Feeds every control agent and the genealogy.
Pick, transfer, orient and stack. Diverts flagged ware on Seedscan's signal.
Simulates the whole chain and ranks candidate recipes. Validates every move before promotion.
The Glasent SDK is typed Python. Tools are declared with schemas and limits; the policy engine enforces them at call time, not in a review meeting.
# Bound the anneal agent to twelve lehr zones on FL-2.
from glasent import Agent, Tool, Limit, Autonomy
lehr = Tool(
name="lehr.write_curve",
tags=["FL2.LEHR.Z01..Z12.TEMP_SP"],
limits=[Limit(max_step="4 C", per="60s")],
)
anneal = Agent(
id="agent.anneal_stress",
goal="residual stress inside spec, min energy",
tools=[lehr, Tool("optic.read_birefringence", read_only=True)],
# bounded writes; a technologist still gates pull steps
autonomy=Autonomy.L3,
# simulate on Glastwin before every write
verify="twin",
)
run = anneal.start(line="FL-2", product="CLR-4MM")
for step in run.stream():
print(step.name, step.status, step.duration)No results are claimed yet. This is the loop each agent closes; the pilot measures the number.
| Agent | Watches | Moves | Tier |
|---|---|---|---|
| Meltrix | Furnace zone temperatures, pull rate, batch weights, redox and fining signals, lab chemistry, Seedscan seed and stone rate | Batch dosing, furnace zone temperature, pull rate, inside the approved composition window | Line or Plant |
| Formeon | Gob weight and temperature, ribbon speed, top-roller angle, tin-bath conditions, thickness and flatness scans, Seedscan check risk | Gob weight, forming timing, ribbon speed and top-roller settings, inside per-tag rate and magnitude limits | Line or Plant |
| Anneon | Lehr zone temperatures, belt speed, cooling rate, birefringence and polariscope readings, product thickness from Formeon | Zone setpoints and the cooling curve, inside the approved stress and energy envelope | Line or Plant |
| Seedscan | Every camera, optical, thermal and stress station on the line, at 30 to 60 frames a second | Nothing directly. It raises typed signals and alarms that the control agents and the orchestrator act on | Line or Plant |
| Panebot | Cell cameras, ware geometry and temperature, Seedscan defect flags per piece, robot and conveyor state | Pick points, grip force, paths and stack destinations, inside the cell's safety envelope | Plant |
| Glastwin | The as-run state of the furnace, forming line and lehr, plus every candidate recipe the orchestrator proposes | Nothing on the plant. It ranks and validates candidate moves and hands the winner to the control agents | Plant |
No customer quotes yet; we are pre-launch. These are the three buyer personas the product is built for and the pain each one describes, in their own terms.
"The furnace has run for years on the same setpoints. The process never has. We find out a melt drifted when the cullet pile grows."
Plant / operations director · ICP persona
"I can tell you why a check appeared from the lehr curve and the gob weight. I cannot be at every line, and the people who could are retiring."
Glass technologist · ICP persona
"A missed seed is a reject. A missed stress fault is a pane that shatters in the field. I need genealogy on every piece, not a spot check."
Quality / reliability engineer · ICP persona
Glasent reads and writes through the furnace, forming, lehr, inspection and MES systems already on the floor. No rip-and-replace, no parallel historian, no new HMI to learn.
Furnace SCADA and PLC, batch-plant weighing, redox and fining instruments
Setpoint reads and guarded writes over OPC UA and Modbus
Float-bath and IS-machine controls, gob-weight and timing systems
Gob, pull, ribbon speed and roller reads; guarded writes
Lehr zone controllers and belt drives
Zone temperature and curve reads; guarded writes
Camera, optical, thermal, polariscope and birefringence stations
Frames, stress maps, defect records, line-speed streams
Orders, grades, lots and ware genealogy
Spec and tolerance reads; genealogy and release writes
Time-series stores and lab information systems
Backfill, replay and lab chemistry
Robot cells, conveyors and stackers via NVIDIA Isaac
Pick, path and stack commands inside the safety envelope
SSO and RBAC via SAML or OIDC; NVIDIA Jetson Orin edge nodes
Named approvers, tag-level roles, sub-100 ms inference
Glasent 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.
| Standard | Scope | Status |
|---|---|---|
| SOC 2 Type I | Cloud control plane | RUNNING Planned in the first six months |
| SOC 2 Type II | Cloud control plane | QUEUED Planned in months six to twelve |
| IEC 62443 | Plant-edge OT security | RUNNING Design-aligned |
| ISO 9001 / IATF 16949 | Quality and genealogy records | SUCCEEDED Record formats supported |
| Container and safety-glass standards | Stress and defect conformance records | SUCCEEDED Record formats supported |
One policy model, one audit trail, one benchmark across every line in every plant, with the composition and forming models kept private to each site.
The questions plant directors and glass technologists 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 site has set. Every pilot starts in shadow mode, where Glasent predicts and recommends and a person enters everything. Writes come later, after the recommendations have earned it.
Control returns to your existing furnace, forming and lehr systems at their last known-good state. Glasent is a supervisory layer on top of the control system you already run, never a replacement for it, so an outage degrades the plant to its current way of running, not to a stop.
Shadow mode starts on the first day from existing SCADA, forming, lehr and inspection data. Defect and stress prediction improve as site history and labelled outcomes accumulate; the pilot plan sets a baseline period before any recommendation is scored.
Only if you choose cloud training. Compositions, forming recipes and defect libraries are tenant-isolated and never used to train another customer's models. On-prem training and an air-gapped plant edge are available for IP-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 90 to 120 day line pilot in three stages: shadow mode to measure the baseline, assist mode where a technologist approves each recommendation, then bounded write-back on low-risk forming, annealing or inspection loops if the plant is satisfied with the results.
Pick the section that hurts: melting, forming, annealing or inspection. One agent, one line, one baseline, 90 to 120 days.
↑↓ navigate↵ openesc close