Wired determinism worked. Wireless broke the assumption.

Deterministic wired networks worked for years. Adding wireless and roaming introduced the hard problems.

That isn’t nostalgia. A wired control network earns its trust from things nobody has to think about. A device stays on the same port. The cable doesn’t move. Cyclic I/O shows up on schedule, and the timeouts in the controller were set with that in mind.

Wireless breaks all three at once. The medium is shared and half-duplex, so every frame waits for a turn, and no station is promised one (RFC 8325¹). The device moves. And when it moves far enough, it roams: it leaves one access point for another, and for that moment it is on no network at all.

The roam is the part I’d underline. Interference comes and goes. Congestion builds slowly enough to see. A roam is an outage the client schedules for itself, on every pass down every aisle, and it lands where the coverage is thinnest.

Here’s the honest part: Wi-Fi keeps getting better at roaming. Fast transition and key caching exist to shrink exactly this gap, and they do. I’d still argue the comparison is the wrong one. “Better” is measured against the last version of Wi-Fi. The plant measures it against the application’s timeout.

Take a CIP Safety input connection at its documented defaults. The controller allows 40 ms² without a valid packet before it faults the connection. Whether a roam fits inside that depends on what is connected, and on whether safety runs over the link at all. Safety over wireless needs very low roam times to avoid safety timeouts.

The cost of missing is not a slow page load. One AGV disconnecting on an assembly line shuts down the entire line.

So the question changes. On a wired network you asked whether the network was up. On a wireless network you ask how long the worst outage lasts, and whether that worst case fits inside the timeout of everything riding on it. That is a translation job, from what the application tolerates to what the wireless design must deliver. It’s the job this site is built around.

Wired determinism worked because nothing moved. On the plant floor now, everything does.

Ben Rutter. "Wired determinism worked. Wireless broke the assumption.." OT Wireless, published October 5, 2026. https://otwireless.com/perspectives/wired-determinism-worked/

Rutter, B. (2026, October 5). Wired determinism worked. Wireless broke the assumption.. OT Wireless. https://otwireless.com/perspectives/wired-determinism-worked/

@misc{rutter2026wireddeterminismworked,
  author = {Rutter, Ben},
  title = {{Wired determinism worked. Wireless broke the assumption.}},
  year = {2026},
  howpublished = {\url{https://otwireless.com/perspectives/wired-determinism-worked/}},
  organization = {OT Wireless},
}