Watch it work.

Break a bottle filling line on purpose — a conveyor that stalls, a nozzle that clogs, a sensor that lies — then test against a machine that fails.

5 parts: build the twin · program the sequence · automated testing · virtual commissioning · scenario testing. Press play, all five run in order.

Start here. Building a machine as states and the conditions that move between them — the drawing that turns out to be the program.

Write the tests, run the suite after every change, and read a coverage report that names the transitions nobody ever tried.

The same control logic, driving a 3D model of the machine. You watch the sequence run, and find the ordering mistakes while the real machine is still on a drawing.

Real EtherCAT I/O driven by StateTick — and what to do about the half of the machine that has no sensors on it yet. The twin stands in for the hardware that has not arrived, so the same logic runs against real drives and simulated ones at once.

Building the operator panel against the controller’s own variables — so the panel and the logic cannot disagree about what anything is called.

Every real machine gets stopped, held and resumed. In StateTick that is behaviour the controller provides, not scaffolding you rebuild around every sequence.