Automation Downtime Explained: What Causes It, What It Costs, and How to Prevent It

Key takeaway: Automation downtime is any period when an automated system stops producing — a robot cell halts, a conveyor jams, or control software crashes. Unplanned downtime is the expensive kind: Siemens puts the average cost at $260,000 per hour across industries, and the world’s five largest companies lose $1.4 trillion a year to it. Most of it is preventable.

What Is Automation Downtime?

Automation downtime is the time an automated system is not running when it should be. It covers factory robot cells, packaging lines, warehouse conveyors, and the software that controls them. The term also applies outside factories — to automated IT systems, data pipelines, and self-service kiosks — but manufacturing is where the cost is measured most carefully.

There are two kinds. Planned downtime is scheduled: a maintenance window, a software update, or a changeover to a new product. It is budgeted, staffed, and short. Unplanned downtime is the opposite: a breakdown, a jam, a power cut, or a cyber incident that stops the line without warning. Everything in this article is about the unplanned kind, because that is where the money goes.

The scale is large. Siemens’ 2024 report The True Cost of Downtime, based on 181 interviews with maintenance, engineering, and IT professionals, found that the average facility suffers 25 downtime incidents a month. That is better than 2019, when the average was 42 — but the cost per incident has climbed. Facilities lose an average of 30 hours of production to downtime each month, and 17 of those hours are unplanned, according to L2L’s downtime research.

The Main Causes of Automation Downtime

Most downtime traces back to a small set of causes. The table below covers the ones maintenance teams report most often.

Cause	What happens	Typical fix
Equipment failure	Motors, bearings, sensors, or drives wear out and stop the line	Condition monitoring and scheduled part replacement
Software and control faults	PLC programs, network links, or MES/ERP integrations fail or desync	Version control, tested updates, redundant networks
Human error	Wrong settings, skipped steps, or undocumented fixes after a repair	Checklists, training, and recording every change
Ageing machinery	Old equipment breaks more often and spares become hard to find	Planned upgrades before parts go obsolete
Supply and utility failures	Missing parts, power cuts, or a supplier delay stops the line	Critical spares on site and backup power
Cyber incidents	Ransomware or a network intrusion forces a shutdown	Segmented networks and tested recovery plans

Equipment failure is still the most common trigger, and ageing machinery makes it worse: plants running equipment past its intended lifecycle see reliability fall while maintenance needs rise. Human error is the quieter cause. L2L found that 72% of manufacturing teams say undocumented fixes — repairs and machine changes nobody records — contribute to downtime. A fix that is not written down is a fault waiting to repeat. And supply chain disruption accounts for roughly 30% of all manufacturing downtime, according to ACI Controls.

Automation Downtime vs Planned Downtime

The difference between the two is not just scheduling — it is cost, control, and consequence. A planned stop happens on the plant’s terms, with parts, people, and a restart plan ready. An unplanned stop happens on the machine’s terms, usually at the worst moment, with an emergency parts order and overtime labour.

The cost gap is wide. A planned maintenance window is measured in labour hours. An unplanned outage averages four hours, according to ACI Controls, and 82% of companies experienced at least one unplanned outage in the past three years. The same research puts the cross-industry average at $260,000 per lost hour, with large automotive plants losing about $2.3 million per hour and oil and gas operations around $1 million per hour. ABB’s Value of Reliability survey of 3,215 plant maintenance leaders put the typical industrial business at $124,669 per hour. Even a conservative figure from Gartner’s 2022 data — $8,581 per minute — turns a four-hour stop into a bill above $2 million.

What Does Automation Downtime Cost?

The bill has four parts. Lost production value is the visible one: products that were never made and cannot be sold. Idle labour is the second: operators and technicians paid while the line stands still. Recovery costs are the third: emergency spares with expedited shipping, specialist contractors, and overtime. Reputation is the fourth and hardest to measure: missed shipments, contractual penalties, and customers who start looking for a second supplier. Siemens estimates the world’s five largest companies lose $1.4 trillion a year to unplanned downtime — about 11% of their total revenue.

For a midsize manufacturer the direct costs alone run into tens of thousands of dollars per hour, with indirect and long-term impact pushing the total into the hundreds of thousands, according to ACI Controls. And the trend is not improving on its own: MaintainX’s State of Industrial Maintenance 2026 survey of 2,234 maintenance and operations leaders found that 79% of teams saw unplanned downtime hold steady or increase over the past year — 45% reported no change and 34% reported an increase. Only 21% saw less of it.

What a Plant Needs Before It Can Cut Downtime

Cutting downtime starts before buying any new tool. The prerequisites are practical, not expensive. First, know the baseline: how many incidents, how long they last, and what caused each one. Without that record, no improvement can be measured. Second, keep critical spares on site — the part that takes three days to arrive is the one that will fail on a Friday night. Third, record every fix: the undocumented-change problem above is free to solve and expensive to ignore. Fourth, schedule maintenance around the data, not the calendar: replace the bearing that vibration data says is failing, not the one the calendar says is due.

The tool that ties this together is predictive maintenance. It uses sensor data and historical patterns to flag a failure before it happens, so a repair becomes a planned stop instead of an unplanned one. Only about 27% of manufacturing facilities used predictive maintenance in 2024, according to MaintainX — but Siemens reports that companies using it have cut unplanned machine downtime by 50%. For a closer look at the upgrade side of this, see our guide to industrial automation upgrades and our explainer on IoT and automation, which covers the sensors that make prediction possible.

The Risks and Limits of Automation Itself

Automation reduces some downtime and creates new kinds of it. A fully automated line has more sensors, more software, and more network dependencies than the manual process it replaced — each one a new failure point. A single PLC fault or a failed software update can stop a line that used to keep running with a spanner and a workaround. Cyber risk grows the same way: connected machines are reachable machines, and the FBI reported ransomware remained the most pervasive threat to US critical infrastructure in 2024.

The practical answer is not less automation but more resilience: redundant networks, tested backups, segmented control systems, and a recovery plan that has actually been rehearsed. A plant that can restart a line from a known-good state in minutes is in a different position from one that discovers its backup was never tested — during an outage.

Analysis: Downtime Is a Data Problem Before It Is a Machine Problem

Analysis: The numbers in this article point the same way. The plants still losing ground are not the ones with the oldest machines — they are the ones with the least information about their machines. The 50% reduction Siemens credits to predictive maintenance and the 72% of teams blaming undocumented fixes describe the same gap from opposite sides: nobody knew the failure was coming, and nobody wrote down the last one. A plant that records every incident, stocks the spares its data says it needs, and fixes the failing bearing on Tuesday instead of the broken machine on Saturday will beat a plant that buys newer robots and changes nothing else. Downtime is mostly a management problem wearing a machine’s overalls.

Frequently Asked Questions

What is automation downtime?

Automation downtime is any period when an automated system stops producing when it should be running — for example, a robot cell halts, a conveyor jams, or control software crashes. Planned downtime is scheduled maintenance; unplanned downtime is a breakdown or fault that stops the line without warning, and it is far more expensive.

How much does automation downtime cost?

The cross-industry average is about $260,000 per hour, according to Siemens’ 2024 True Cost of Downtime report. Large automotive plants lose around $2.3 million per hour and oil and gas operations about $1 million per hour, while ABB’s survey of 3,215 plant leaders put the typical industrial business at $124,669 per hour. The total includes lost production, idle labour, recovery costs, and reputational damage.

What causes most automation downtime?

Equipment failure is the most common cause, made worse by ageing machinery and missing spares. Software and control faults, human error — especially undocumented fixes, which 72% of manufacturing teams say contribute to downtime — supply chain disruption, and cyber incidents are the other main causes. Supply chain disruption alone accounts for roughly 30% of manufacturing downtime.

How can a plant reduce automation downtime?

Start by recording every incident, its duration, and its cause, then keep critical spares on site and document every repair. Predictive maintenance — using sensor data to flag failures before they happen — is the most effective step: Siemens reports companies using it cut unplanned machine downtime by 50%. Scheduled upgrades before parts go obsolete also prevent long, expensive stops.

Is planned downtime the same as automation downtime?

No. Planned downtime is a scheduled stop for maintenance, updates, or changeovers, done on the plant’s terms with parts and staff ready. Automation downtime usually means unplanned downtime — a breakdown that stops production without warning. The average unplanned outage lasts about four hours, and 82% of companies experienced at least one in the past three years.