When the factory stops
Sep 29th, 2025 by Jesper Kråkhede
At the beginning of this month a large car manufacturer was hit by a cyberattack and shut down its systems to contain it. Production at its plants stopped. As I write this it has been almost four weeks, the restart has been pushed back several times, and the government has stepped in with a loan guarantee to keep the suppliers afloat.
What strikes me is that, as far as has been reported, the attackers didn´t need to touch the machines on the factory floor. A modern car plant runs on IT: production planning, parts ordering, logistics, supplier portals and manufacturing execution systems. When those go down, the line stops, even if every robot is fine. The boundary between IT and OT isn´t where most people think it is.
What can manufacturers do technically? Map the dependencies. Which IT systems does production actually need to run, and for how long can it run without them? Separate them. Production critical systems should be in their own zone, with their own identities and admin paths, so an attack on the office network doesn´t automatically reach them. Plan for running in isolation. Can the plant keep producing for a few days with the corporate network disconnected? Back up configurations and production data, offline, and test the restore. And practise the shutdown and the restart, because a controlled restart of a whole plant is a complex project in itself.
Just in time production has no buffer. Neither does just in time security.