Fleet control · CNC production

Every machine, every plant, live — and in control.

Undrd Control shows the state of your whole CNC fleet as it happens: which machines are online, which are running, what program is loaded, which tool is wearing, and what needs you now. Beside each machine, an edge agent measures the part and corrects the tool offset before the drift becomes scrap — retrofitted onto the controllers you already own.

See the fleet viewTalk to an engineer

Works with CNC controllers from

  • GSK
  • Fanuc
  • Siemens
  • Heidenhain
  • Fagor
  • Mitsubishi Electric
  • Haas
  • Mazak
  • Okuma
Fleet overview
LiveIllustrative example

Fleet yield

99.2%

Parts today

4,812

Machines online

10 / 12

Open alerts

1

Plant ANominal
99.6%
5 machines
Plant BWarning
97.8%
4 machines
Plant CNominal
99.4%
3 machines
  • Lathe 03 · correction −0.013 mm on T02
  • Lathe 07 · late — last heard 40 s ago
  • Lathe 07 · back online

Fleet state

Know the state of every machine without walking the floor

Undrd Control is the one place your whole fleet reports to. Every edge agent sends a heartbeat and pushes each change as it happens, so for every machine, in every plant, the dashboard can tell you whether it is online, what it is doing, and whether that needs you. The loop itself runs on the machine; the dashboard is where you see it.

  1. 01

    Fleet

    Yield, parts made, machines online and open alerts across every plant on one screen. Each plant carries a status, so the one that needs you stands out.

  2. 02

    Plant

    Every machine in the plant with its link state, run state and active program, plus performance and defect-rate trends you can zoom into.

  3. 03

    Machine

    Spindle and feed, tool wear against its limit, active alarms with their manual text, and the program that is really running.

Fleet overviewIllustrative example
Plant ANominal99.6%
Plant BWarning97.8%
Plant CNominal99.4%
Plant B · machinesIllustrative example
  • Lathe 05Online
  • Lathe 06Online
  • Lathe 07Late
  • Lathe 08Offline
Plant A · Lathe 02Illustrative example
Online
Run state
Running
Program
O1003
Spindle / feed
2 400 rpm · 120 mm/min
Active alarms
None

Tool wear vs limit · mm

  • T010.041
  • T020.078
  • T030.012

Twin T05 ready

Four link states, always clear

  • Online

    The agent was heard within its heartbeat interval. Readings are current.

  • Late

    A heartbeat was missed. Shown in amber with when it was last heard, so a brief delay is easy to tell from an outage.

  • Offline

    Silent past the offline window. The last readings stay on screen, dimmed and time-stamped, and a fleet alert fires after 10 minutes by default.

  • CNC unreachable

    The agent is online but the controller is not answering — so you know straight away whether to check the network or the machine.

What you get

  • Live machine status

    A change on the CNC reaches the dashboard within seconds, and a keepalive confirms the link every 15 seconds. Readings dim as they age, so you always know how fresh they are.

  • Alerts that watch the fleet

    Offline machines, stalled commands and growing upload backlogs, alongside defect rate and process drift over a rolling 30 minutes. Acknowledge them, or get them by email.

  • The program that is really running

    The active program is recognised by its content, not its file name, so you always know it is exactly the one you delivered.

  • Alarms, explained

    Each active controller alarm is listed with its description from the machine manual where one exists, so the operator knows straight away what is going on.

  • Tool wear at a glance

    Per-tool wear against its limit, twin tools ready to take over, and a wear log that starts again when the program changes.

  • Access and audit

    Four roles scoped from organisation to plant to machine, an audit log of who changed what, and machines archived with their full quality record.

Preview

Watch a program travel from your desk to the machine

Six steps take a part program from your CAM export to a running, self-correcting cycle. We have laid it out screen by screen, using the dashboard's own views.

  • Upload the program and see what the machine already holds
  • Describe the operations and the probe cadence once
  • Send it with one click and follow the delivery live
Walk through the workflow
01You

Add the program to the machine's library

Upload new referenceIllustrative example

NC file

required

✓shaft_A_rev2.nc

STEP file

optional, for 3D preview

✓shaft_A.step

Upload

Library

O1003

shaft_A_rev2.nc

version 2 of 2

How it works

Measure, correct, observe.

One loop, running on the machine itself. The dashboard is where you follow it.

Data flow

Probemeasuring machineCamera + sensorsprocess signalsEdge agentbeside the machineCNC controlleroffset appliedUndrd Controlfleet dashboarddimensionframes + sensorscorrectiondurable outbox
  1. 01On the machine

    Measure

    A probe or measuring machine reports the finished dimension as soon as the cut ends. In the same moment the camera frame and the process sensors go to the inspection model.

  2. 02In the loop

    Correct

    The edge agent computes the wear correction, clamps it against the per-tool limit, and writes it to the controller before the next part is cut.

  3. 03Wherever you are

    Observe

    Every inspection, correction and alert replicates up to the dashboard. The edge keeps a durable outbox, so if the link drops, data waits on the device and follows as soon as it returns, with nothing lost.

One correction, to scale
next part−0.0300+0.030Ø deviation · mm

Nominal

Ø40.000

Measured

Ø40.026

Tolerance

±0.030

Correction written

−0.013

radius, X wear

Measured on the machine. Corrected before the next part is cut.

Only the edge agent sits on your network, and it always connects outward, so your network stays closed to the outside. The service runs in the cloud, and the agent is built for NAT, outages and reboots.

Capabilities

What we build

Six things we do. Most plants start with the fleet view and one machine, and add the rest over time.

01

Fleet state and quality analytics

Every machine's state, in every plant, on one screen — scoped to what each person is allowed to see.

  • Online, late or offline for every machine from its own heartbeat, with readings that dim when they go stale and a separate state for a controller that stops answering
  • Organisations, plants and devices, with separate scopes for system administrators, senior managers, managers and operators
More detail · 5
  • A rolling alert engine over a 30-minute window, with acknowledgement and delivery by email or in-app
  • Fleet alerts too: a device that goes offline, a command that stalls, an edge agent whose upload backlog is growing
  • Quality events with a formal disposition, and an audit log of who changed what
  • Throughput, pass rate and defect breakdown — per device, per plant, across the fleet
  • English and Portuguese throughout
02

Closed-loop tool-wear correction

The part is measured, the offset is recalculated, and the machine is corrected — automatically, every cycle.

  • Reads the finished diameter straight from the measuring machine or an in-process probe
  • Computes the radius correction, then applies a deadband, a per-cycle clamp and a hard per-tool wear limit before every write
More detail · 3
  • Writes the offset back as a G10 block over RS-232, or straight into a Modbus holding register
  • When a tool reaches its wear limit, the agent swaps in its twin tool by rewriting the running program, then resets the wear count
  • Every correction is logged per tool, so wear becomes a trend you can watch and plan for
03

NC programs and toolpath verification

Manage the programs your machines actually run, and check them before they cut.

  • A per-device program store with operation metadata, probe links and per-tool enable or disable
  • Send a program to the controller from the dashboard, and see which one is loaded right now — matched by content, not by file name
More detail · 3
  • Set how often the probe runs (every Nth part); the dashboard adds the probing cadence and the result hand-off to the program before it is sent
  • Drop a STEP file and its G-code into the browser to see the part and the toolpath together in 3D, with probe moves animated
  • Work-offset inputs, so you can check the program against your actual setup
04

AI part inspection

A multi-modal model that judges the image and the process signature at the same time.

  • A vision branch reads the camera frame while a second branch reads temperature, pressure, vibration, cycle time and dimensional deviation — both fused into a single verdict
  • Also catches the part that photographs clean but ran outside its process window
More detail · 2
  • Every prediction reaches the dashboard with its defect class and confidence
  • Defect frames are kept, so an operator can see exactly what the model saw
05

Legacy machine connectivity

Bring machines built before "Industry 4.0" was a phrase onto the network — without replacing them.

  • Serial, fieldbus or a vendor API — the first job is working out what your controller will actually expose, because that varies by model and by vintage
  • Every controller sits behind one fixed interface, so a machine we have not met before is just one more adapter
More detail · 4
  • Reads measurements out and writes corrections back on the same link, with the protocol and timing matched to the machine
  • Lists, sends and reads back NC programs on the controller itself where the link allows it, not just wear values
  • Where a controller offers no direct way in, a relay driven from the edge unit triggers external cycle start
  • The edge agent runs on a Raspberry Pi or Revolution Pi beside the machine and keeps working even without a network
06

Safety and piece watchdogs

Vision watchdogs that stop the line for what a fixture sensor cannot see.

  • Person detection inside a configured zone signals STOP, and RESUME once the zone has been clear for a set number of frames
  • A three-class piece watchdog separates absent, correctly seated and incorrectly seated parts before the cycle starts
More detail · 2
  • Hysteresis on every signal, so the line stops only on a confirmed detection
  • Deployed alongside your existing interlocks, as an extra layer

! This is a software watchdog. It complements certified safety hardware (ISO 13849 / IEC 61508) and must never be the only safety mechanism on a machine.

The edge agent

A compact unit beside the machine that runs on its own

The edge agent measures, corrects and reports from the shop floor. It is built for factory networks: it connects outward only, keeps working when the internet is down, and writes only corrections that fall within the limits you set.

01

Secure by design

The agent talks to the dashboard over HTTPS, always on its own initiative, so your network stays closed to the outside. Work sent from the dashboard arrives on a connection held open for it, typically within about a second.

02

Rides out outages

Corrections are saved to disk before they are sent, and uploads go through a durable outbox. When the link comes back everything follows in order, and each command is applied exactly once.

03

Checks before it writes

Every correction passes a deadband, a per-cycle clamp and the tool's wear limit. On a GSK 988TA with an in-process probe, it also confirms that the program on the control is exactly the one it delivered before correcting.

04

Twin-tool swap

At a tool's wear limit, the agent rewrites the running program to use its twin, loads it on the control and starts the wear count again — all without a trip to the machine.

05

Its own identity

Each device has its own ID and key, generated and rotated from the dashboard and stored hashed on the server. Retire one device while the rest carry on.

06

Watches itself

The heartbeat reports uptime, temperature and backlog. It ships as a container image that starts at boot and updates itself from the registry, with structured logs for when you need to look inside.

Rule-based vs learned inspection

A rule checks what someone thought to write down. A model learns from what actually goes wrong.

Most factory inspection still runs on thresholds somebody tuned by hand. That solves what it was built to solve: a known feature, in a known place, measured the same way on every part. What it cannot do is adapt when the failure looks different from the one the rule was written for — and it only looks once the part is finished.

Industry 3.0

Rule-based, end of line

  • Hand-coded if/then thresholds, written once per defect type
  • Re-tuning every time the lighting, fixture or part orientation shifts
  • Blind to any defect nobody thought to write a rule for
  • Judges finished parts, so by the time it rejects one the material is already cut
Industry 4.0

Learned, in-process

  • One model learns defects from labelled examples of your own parts, instead of from a rule written per defect type
  • Image and process signals judged together, so a part that looks clean but ran outside its process window is still caught
  • A new defect class is a retraining job on new examples, not a rewrite of the rule set
  • Runs while the part is being made, so drift is corrected before it becomes scrap

Integration

Fitted to the machines you run

Brands, controllers and vintages differ from one floor to the next, and so does what a machine will actually let you read. So we start from what you have: we work out what each controller can expose and build the integration around it.

How a machine gets connected

  1. 01

    Survey what is on the floor

    Which controllers are on the machines, what each one exposes, what is already wired, and how a part gets measured today. This is the step that decides everything after it, and it is where most of the surprises live.

  2. 02

    Build the integration for it

    Every controller sits behind the same fixed interface, so bringing in one we have not met before is a well-defined piece of work. Protocol, signals and timing get matched to the machine in front of us.

  3. 03

    Close the loop on your line

    Inspection, correction and reporting run against your tolerances and your cycle time. What the system does when a part drifts is a decision we make together.

Running another controller? It is one more adapter behind the same interface, not a new product.

Deployment

  • Edge agent on a Raspberry Pi or Revolution Pi, mounted beside the machine and updated automatically from a container registry
  • Undrd Control is delivered as a managed cloud service, so the servers are ours to install, patch and run
  • What leaves the plant is agreed with you before the system goes live — your production data stays yours
  • Multi-instance from the start: one agent per machine, one dashboard for the fleet

CNC controllers supported today

  • GSK 980TD
  • GSK 980TC3
  • GSK 988TA / 988TA1 / 988TB
  • Fanuc 0i series
  • Fanuc 30i series
  • Fanuc with Data Server board

Impact

What Quality 4.0 is worth

These are published benchmarks for the category, presented as such. What your line returns depends on your scrap rate, your cycle time, and how much of your inspection is still done by hand.

30–50%

Reduction in machine downtime

attributed to predictive maintenance in smart-factory deployments

McKinsey

10–20%

Reduction in cost of quality

across supply-chain technology programmes

Gartner

In-process

Not end-of-line

the shift from inspecting finished parts to correcting the process that makes them

IBM

Figures are industry benchmarks published by the sources named, not measured results from Undrd installations.

Custom work

Custom development for what your factory actually does

No two shop floors are the same, and most of the interesting work sits in the gap between a platform and a plant. The system is built to be extended to fit your line.

New machine adapters

Controllers and measuring machines beyond the supported list, written against the same interface as the rest.

Models trained on your parts

Inspection models built on your defects — including one-class setups for lines where you only have images of good parts.

Line-specific logic

Gauging routines, sorting gates, traceability, and integration with the MES or ERP you already run.

Retrofit engineering

Wiring, I/O, edge hardware and the commissioning it takes to run reliably on a real floor.

Undrd Control is delivered as a managed service, so the infrastructure is ours to run. What gets collected, where it is kept and who can see it are settled with you before the system goes live — keeping your production data private is part of the engagement from day one.

Questions

What people ask first

How do I know a machine has stopped reporting?

Its card turns amber as soon as a heartbeat is missed and red once it passes the offline window, with the time it was last heard. If it stays offline, a fleet alert fires — by default after 10 minutes — in the dashboard and by email.

Do I have to replace my machines?

No. Everything is fitted by retrofit: an edge unit beside the machine, and a link to the controller you already have. The machine keeps running its own program.

Does anything connect into our network?

No. The edge agent always connects outward to the dashboard over HTTPS, so your network stays closed: no ports to open, no firewall changes.

What happens when the internet drops?

The agent keeps measuring and correcting, because all of that happens on the machine itself. Data waits in a durable outbox on the device and is sent as soon as the link returns.

What if our controller is not on the list?

Tell us which one. We look at what it can expose, then write an adapter against the same interface as the others — a well-defined piece of work.

Who can see our production data?

Access is scoped from organisation to plant to device, by role, and every change is in an audit log. What leaves the plant is agreed with you before the system goes live.

Get in touch

Tell us about the line

The useful first conversation is a short one: what you make, what you scrap, and which controllers are on the floor. That is usually enough to see how we can help.

Send us the details[email protected]