OPEN
Find the bottleneck
Open the plant overview and identify the asset with production impact.
Short demo · 5 minutes · Technical tour · about 28 minutes
Experience ZenFactory step by step
This page has two entry points: a short decision story for the first impression and the full technical tour for the review afterwards. Data path, Policy Gate, human approval, trial run and evidence stay visible throughout.
Short demo Five minutes to the result
for first contact, management and procurement- 01 Understand the situation 1 min 1,000 parts by 5 pm, tool wear at ST020 and a 70 kW limit. The demo starts with the trade-off, not a diagram. open →
- 02 Choose an option 1 min Reduce feed, shift part of the order, or change nothing: each option shows delivery, energy, tool life and risk side by side. open →
- 03 Who is in control 1 min Policy Gate and human approval stand in front of every write path. The software makes proposals, not covert machine commands. open →
- 04 Measure the effect 1 min Canary, watchdog and expected-vs-actual show whether the predicted effect is larger than the metric's noise. open →
- 05 Take the evidence 1 min The compact outcome report summarises goal, decision, effect, rollback and sources so it can be shared internally. open →
The short demo shows a reproducible scenario. The plant runs sped up;
the protocols, database, asset administration shell, policy check and evidence chain are
the vitrine's real system paths.
Technical tour
0/9
Progress stays local in this browser. The demo does not write anything to plants because of it.
Show now · step 01
The factory is running
Plant overview: five plants, two sites, all KPIs live.
What to say?
Five plant types under one model — a chained production line, a cell with five parallel machines, a filling line with collection tables, two batch reactors and a band saw from 1998 that only speaks Modbus. Each computes its own physics.
What does it prove?
This is not a backdrop with random numbers. Every number comes from a model that you can question and recompute.
Interactive process · Industry 5.0 loop
Don’t read it — play it through
The animated packets show that data flows outward, while every path back through OT runs via Policy Gate, human approval, Canary and Watchdog. Click a phase and the metrics change with it.
01 OPC UA sample 02 Sparkplug B 03 AAS twin 04 Policy Gate 05 Human approval 06 Canary window
MQTT
CONNECTED
Historian
WRITING
AAS
UPDATED
Policy
ARMED
Watchdog
ACTIVE
Ledger
SIGNED
Active state
ST020 reports tool wear
Vibration, spindle temperature and RUL flow into the semantic twin.
RUL 47 ± 8 min
Vibration +18 %
Power cap 70 kW
Mission Mode · technical challenge
Six tasks that make the demo verifiable
Progress
0/6
The status stays local in this browser. Nothing is written to the plant.
OPEN
Check the data path
Follow one measured value from shopfloor through edge, OT-DMZ and historian.
OPEN
Evaluate an option
Compare reducing feed, moving work and making no change.
OPEN
Read the Policy Gate
Check why Safety, Quality and Energy are evaluated separately.
OPEN
Trigger resilience
Simulate a disturbance and inspect buffering, backfill and data loss.
OPEN
Replay the audit
Reconstruct trigger, approval, execution and measured effect.
Experience Industry 5.0 · Augsburg plant
From event to proven decision
Order: 1000 parts by 17:00. Initial state: OEE 84 %, energy forecast 182 kWh, CO₂ 71 kg, scrap 1.1 %. Then ST020 reports declining remaining tool life and the plant activates a 70 kW power limit.
Event 1
ST020 predicted tool degradation
RUL 47 ± 8 min · confidence 0.87
Event 2
Energy constraint activated
Plant power shall remain below 70 kW
Evidence
vibration RMS +18 % · spindle temp +7 %
classical analytics + policy threshold
Authority
AI creates proposals
Policies and safety logic remain authoritative
Alternatives instead of Approve/Reject
Three action options with trade-offs
Option A
Reduce ST020 feed by 8 %
Throughput -3,1 %
Peak Energy -11,8 %
Tool lifetime +14 %
Scrap risk -0,3 pp
The delivery deadline remains reachable, but the buffer until 17:00 shrinks.
Option B
Move part of the order to ST030
Throughput +1,8 %
Peak Energy +0,8 %
Tool lifetime ST020 +18 %
ST030 utilization +9 %
Better delivery performance, but higher utilization and more energy in the secondary path.
Option C
No change
Delivery risk +14 %
Peak Energy unchanged
Tool failure probability +22 %
Operator workload +1
No intervention, but rising risk of an unplanned tool change.
Policy Gate
Safetyonly approved setpoint limits PASS
Qualityscrap risk decreases in the canary window PASS
Energy70 kW limit is maintained PASS
Human approvalapproval required because production is affected WAITING
Controlled Execution
- 1Human Approval
- 2Canary
- 3Watchdog
- 4Execute
- 5Measure
- 6Validate
- 7Audit Ledger
Expected vs Actual · Decision verified
| Metric | Expected | Actual |
|---|---|---|
| Peak Energy | −18,0 % | −16,9 % |
| Scrap | −0.40 pp | −0.35 pp |
| OEE | −1,2 % | −1,4 % |
| Delivery | on time | on time |
Audit Event DEC-2026-000381: Trigger Tool degradation, Policy PASS, Human APPROVED, Execution SUCCESS, Rollback READY. In the showcase this is a reproducible scenario; in a real plant the same controlled write path remains in place.
Outcome Report
One decision, ready to share internally
Goal
1,000 parts by 5 pm, under 70 kW, no loss of quality
Decision
ST020 feed −8%, human approval, canary rather than immediate adoption
Result
Peak energy −16.9%, scrap −0.35 pp, delivery on time
Uncertainty
OEE −1.4%, within the expected band; small effects remain flagged as noise
Control
Watchdog active, rollback ready, no direct AI write access
Sources
AAS snapshot, historian window, Policy Gate, audit ledger
Audit Replay · reconstruct decision
Decision ID
DEC-2026-000381
Trigger
Tool degradation ST020 + energy constraint
Observations
vibration RMS +18 % · spindle temp +7 % · RUL 47 ± 8 min
Proposal
feed −8 %, partial move to ST030, tool change window 14:45
Policy evaluation
Safety PASS · Quality PASS · Energy PASS
Human decision
APPROVED
Execution
Canary → Watchdog → Execute → Measure → Validate
Actual result
Peak Energy −16.9 % · Scrap −0.35 pp · delivery on time
Rollback state
READY
OT Security Architect Review
Architecture review instead of just an architecture diagram
Direct IT → OT access
DENIEDno IT access to controllers
Broker / historian-facing
OT-DMZseparated from the control network
AI direct write
NOAI creates proposals, not commands
Policy before write-back
YESdeterministic before Human Approval
Canary / Watchdog
READYcontrolled trial run with abort
Audit chain
INTACTHash-chained and reconstructable
open proposals 0 For step 6, a proposal must be open. The loop creates one itself as soon as
the situation warrants it.
- 01
The factory is running
2 min Go thereShow
Plant overview: five plants, two sites, all KPIs live.
Say
Five plant types under one model — a chained production line, a cell with five parallel machines, a filling line with collection tables, two batch reactors and a band saw from 1998 that only speaks Modbus. Each computes its own physics.
Proves
This is not a backdrop with random numbers. Every number comes from a model that you can question and recompute.
- 02
One number, its whole journey
3 min Go thereShow
Plant page, hall view, then click a station.
Say
This station’s cycle time comes from the simulation, is offered over OPC UA, published by the edge gateway as Sparkplug B, written into TimescaleDB and kept in the digital twin under a standardised path. Five stations, no step skipped.
Proves
The chain is complete and carries real protocols — not a REST interface pretending to.
- 03
The machine that cannot do OPC UA
3 min Go thereShow
Cutting: the same plant page as before, the same KPIs — and below it the note that here things are only read.
Say
This band saw is from 1998. It speaks no OPC UA and never will; it has ten holding registers and an indicator lamp. What you see is calculated by a connector at its side from four counters and an operating mode — availability, performance, quality, OEE. In the AAS tree, in the historian and in this interface it is indistinguishable from the other four plants.
Proves
The most common objection is “my machines can’t do that”. They don’t have to. What they have to be able to do is count — and any machine can. What it cannot do is stated there too: this plant has no intervention button, because there is no write path.
- 04
Trigger a fault
3 min Go thereShow
The button here triggers a tool breakage at ST020. Then switch to the plant page. In the public showcase this is visible as a recorded state.
Say
The call goes through the OPC UA method InjectFault — the same way a third-party control system would have to go. Watch the buffer in front fill up and the stations behind starve.
Proves
The layers really are separate. And: buffers decouple — output collapses later than the station does, and that is exactly why the OEE of a line is not a simple multiplication.
- 05
What the control loop makes of it
4 min Go thereShow
Control-loop page, go through all four tabs.
Say
Four plants, one control loop. What differs is the domain knowledge: here the bottleneck, there the queue, there the back-pressure, there the exotherm. Every proposal states its reason, its assumption and the expected effect in numbers.
Proves
The expected effect is fixed beforehand. Without it, it could not be checked afterwards whether the proposal was good — you would have automation without learning.
- 06
Approve a proposal
4 min Go thereShow
Approve an open proposal and watch the trial run.
Say
Approved does not mean adopted. The value is run for a limited time, and afterwards the measured effect decides. The watchdog runs alongside the whole time and aborts without waiting for the end.
Proves
The difficult part is not the intervention but the proof that the forecasts are right. And that takes time: a trial run lasts two to six hours of plant time so that the effect stands out from the noise. At 20× time-lapse that is six to eighteen minutes — start it here and come back to it at step 9.
- 07
The evidence
3 min Go thereShow
Chain of evidence at the bottom of the page, plus the check at the top right.
Say
Every step is listed here: who, what, why, through which gate, with whose approval. Hash-chained and signed. The database rejects UPDATE, DELETE and TRUNCATE — whoever changes an entry breaks the chain from that point on, and the check names the number.
Proves
That is the argument a dashboard cannot make. And the precondition for an operator ever giving the software more freedom.
- 08
Looking back
2 min Go thereShow
Time travel: query the state at any point in time.
Say
No value is ever overwritten. The twin from two hours ago can be queried just like the one from now — and with it the situation on which a decision was based.
Proves
A decision can be reconstructed against exactly the state on which it was made. Without that, an evidence ledger is just a claim with a timestamp.
- 09
Why the optimum is not at the limit
4 min Go thereShow
Control-loop page and Industry 5.0 scenario: options, trade-offs and measured effect.
Say
In the reactor, the number of batches keeps rising up to 92 degrees. The number of saleable batches peaks at 88 and then collapses. Same runs, two answers. Then comes the part almost nobody shows: Policy Gate, canary run and Expected vs Actual show which effect can actually be proven.
Proves
Whoever optimises the wrong KPI drives into an emergency shutdown — which is why an optimiser without domain knowledge of the plant is dangerous. And whoever claims an effect smaller than the noise of their KPI claims nothing at all. The loop then says “indistinguishable” instead of “confirmed”.
What to expect
- “Does it really run?” — Yes. The plant runs at 20× speed so that a shift can be shown in a quarter of an hour. Everything else is real: protocols, database, twin, evidence.
- “Why is the loop proposing nothing right now?” — Because the situation does not warrant it. A loop that adjusts something at every opportunity is not one. The control-loop page gives the reason in plain language.
- “And if the AI talks nonsense?” — It may do nothing without approval. After that it runs for a limited time, is measured, and goes back if the effect fails to appear. Autonomy is off, and the policy states what would have to happen for it to be up for debate.
- “Why does it say for the cell that nothing is measurable?” — Because it is true. There the manipulated variable acts on one of five parallel machines, and the effect stays below the natural fluctuation. Writing that down openly is the difference between a measurement and a claim — and the most convincing argument of this demo.