Industrial Automation

From the machine to the management report.

Four products that cover a plant end to end: one that runs the line, one that reads the machines, one that runs the orders, and one that prints the labels. They are in daily production across multi-site manufacturing groups, and each of them works on its own if that is where you need to start.

Book a demo Get a briefing
In daily production across multi-site manufacturing groups
OPC-UASiemens S7RESTCSVOn-premiseAir-gappedMulti-siteShop-floor screens

Four products, one chain

Signals become states, states become orders, orders become the report.

Machine control

Octopus →

Drives labellers, palletizers, wrappers, strappers and vision checks as one coordinated line. A recipe change reaches every machine that needs it, and every plant runs the same software with its own line profile.

  • Talks to Siemens S7 controllers directly
  • A new machine type is a module, not a new version
  • Operator screens in six languages

Machine data

Panda →

Reads your machines, whatever they speak, and turns the signals into numbers people can act on: output, downtime, speed losses, quality stops. Runs on a small appliance in the plant so it keeps recording when the network does not.

  • OPC-UA, Siemens S7, ADAM, REST, plain TCP and CSV
  • Derived values you define once, calculated for you
  • Dashboards ready on the day it is installed

Manufacturing execution

Panda-BDE →

Takes a production order from release to completion: what to run, on which line, in what order. Operators book progress on screens that update live, and supervisors see the plan against reality.

  • Every unit keeps a record of where it ran
  • Every stop gets a reason that compares across plants
  • Finished quantities go back to the ERP without re-keying

Labelling

LabelBee →

Designs the labels production prints and gets the right one to the right printer. Text, barcodes and images are dragged onto a canvas, the fields are bound to live order and pallet values, and you see the finished label with real values before any stock is used.

  • The people who own the label can change it
  • Batch, expiry, SSCC and quantities filled automatically
  • A user manual in six languages ships with it

The whole picture

Five layers, and you can start at any one of them.

How a connected plant fits together Five layers. On the floor sit Octopus, which drives the line as one system, and Panda, which reads every machine and sensor: both talk to the machines directly. What Panda reads lands on one shared stream, kept in a historian alongside it. Panda-BDE runs the orders from that stream and LabelBee prints what goes with them. Your ERP plans into the same picture, alongside Fleet Dispatcher and Plan Studio, both still in development. Above everything sit KPI Studio and Agent Manager, which report on and automate the plant and the planning systems alike. DECIDING Reporting and automation KPI Studio One agreed set of figures, drilled from the year down to a shift. Agent Manager Automates the work around it, in production and in the office. The people who decide Supervisors, planners and the management report PLANNING Deciding what runs next IN DEVELOPMENT Fleet Dispatcher Turns the day's orders into tours: vehicle, driver, site and sequence. IN DEVELOPMENT Plan Studio Plans a production facility against everything a real day throws at it. Your ERP What to make, in what order, by when RUNNING THE ORDER Transaction systems, MES Panda-BDE Runs an order from release to done, booked against what really happened. LabelBee The label and the papers that go with what was just made. Operators on the floor Screens that show the plan against what is actually happening ONE SHARED STREAM Unified namespace One shared stream Everything read off the floor lands here as it happens, ready for anything above to use. Panda-BDE, LabelBee, KPI Studio and your ERP all work from these same numbers. HISTORIAN Every reading kept ON THE FLOOR Machines and control Octopus Drives the line as one system: machines, robot cells, driverless transport and the warehouse. Panda Reads every machine and sensor, whatever it speaks, and puts what it reads onto the stream. What they talk to Controllers, robots, vision checks, print stations, counters and meters Readings and progress travel up Orders and instructions travel down Each one works on its own. Most plants start with one layer, usually reading the machines, and add the next when the first has paid for itself. They are built to fit together, so nothing has to be replaced to take the next step.

How it runs

Each plant gets its own installation, usually on a small appliance that sits in the plant and needs no connection to the outside world. Sites with their own cluster can run the same software there instead. Nothing depends on a link to headquarters staying up, which is the point: the line keeps running and the recording keeps going.

Appliance per plantPrivate KubernetesAir-gapped installLocal certificatesMulti-site rollout24/7 operation

Works with the rest of the range

The plant data does not have to stay in the plant. The same figures feed a group-level reporting cockpit, and an automation can watch the live state and raise the right person when a line drifts off plan.

Enterprise AI Data-Insights KPI Studio Agent Manager Compliance Studio

How to start

Start with one line, not one plant.

We instrument a single line, show you the numbers it produces after a week, and then decide together what the rollout to the rest of the plant should look like.

Book a demo Get a briefing