If you run networks or IT for a control room, you will already know latency is a problem. What you may not appreciate is how fast a few seconds of delay can cascade into safety incidents, huge downtime bills and regulatory headaches.
What’s happening right now (and why it matters)
Downtime is dramatically expensive — and latency lengthens it.
Siemens’ True Cost of Downtime shows an hour of lost production in large facilities ranges from tens of thousands to $2.3M/hr in automotive-scale sites; aggregated global losses run into the trillions. If latency increases Mean Time To Detect (MTTD) or Mean Time To Repair (MTTR), those dollar-per-hour losses grow directly.
Cyber incidents are surging — and they often show up as slowdowns first.
Industrial ransomware events rose sharply (Dragos documents an ~87% increase year-over-year for industrial organizations in 2024), and utilities saw roughly a 70% surge in attacks in 2024 versus 2023. Attackers frequently throttle or disrupt OT traffic before full takeover; causing latency that degrades operator decision-making.
Human performance collapses as latency grows.
Teleoperation and human factors research demonstrates that delays, jitter and packet loss measurably reduce operator accuracy and increase cognitive load; slow feedback causes poorer control decisions and higher operator fatigue. In other words: lag → mistakes.
Questions for IT and Network decision-makers
- How often do we measure one-way latency and jitter from field sensor → control-room display? Do we have SLOs?
- Have we quantified the dollar (and safety) impact of an extra 2–5 seconds in our critical alarm paths?
- Which parts of our secure network stack (VPNs, inspection appliances, cloud gateways) introduce variable latency? Have we stress-tested them during failure modes?
- Are we running decision logic at the edge where milliseconds matter? If not, why?
- Can our operator displays surface latency/confidence signals so operators know whether data is fresh or delayed?
