Unplanned downtime costs the world's 500 largest companies $1.4 trillion a year, or 11% of revenue.
Across industrial sectors generally, downtime runs about $125,000/hour (ABB Value of Reliability Survey, 2023). In distribution and sortation specifically, it's $10,000-$50,000/hour, plus $15-$50 per shipment once SLA cutoffs are missed.
The cost of a failure isn't the failure itself—it's the delay before anyone notices. That delay is called detection latency.
Failures don't start suddenly. They start as small, detectable signals—heat, vibration, drift—days or hours before a stoppage.
Where an operation sits on the reliability maturity curve (reactive → alarm-based → condition-aware → continuous condition intelligence) determines whether a failure gets caught as a work order or lived through as a shift-ending event.
Take the two-minute Reliability Maturity Assessment to see where your operation sits.
The cost of a failure isn't rooted in the failure itself. It's in the delay before anyone notices.
Failures in high-volume logistics operations don't start suddenly. They start as small, detectable signals—a temperature that's crept a few degrees too high, a vibration pattern that's shifted, an electrical connection running warmer than it should. The signal is there well before the stoppage. The problem is that most operations aren't set up to see it until it's already a failure.
That delay has a name: detection latency. And it's expensive, specifically and measurably.
Across industries, unplanned downtime now costs the world's 500 largest companies an estimated $1.4 trillion a year, or roughly 11% of their revenue. The cross-sector median for a single hour of downtime is running near $125,000, according to ABB's Value of Reliability Survey, based on responses from 3,215 plant maintenance decision-makers globally.
That's the headline number. Here's the one that matters more if you run a distribution center or a sortation hub: In these environments specifically, downtime runs $10,000 to $50,000 an hour, plus another $15 to $50 per affected shipment once SLA cutoffs start getting missed.
Two different numbers, one cause. The gap between them is the maturity gap: The cost of a failure scales almost directly with how long it takes an operation to find out something is wrong.
Failures don't begin at the moment a line stops. They begin days or hours earlier, as a chain of small changes that compound.
Take a bearing. Lubrication degrades. Friction rises. Heat builds. Belt alignment shifts in response. Motor load increases to compensate. By the time any of this shows up as a stoppage, the failure has already been developing for a while. It just wasn't visible to anyone watching for it in the usual ways.
This kind of progression shows up across a handful of common categories: bearing wear, VFD enclosure heat, belt tracking drift, induction timing drift, and rising control cabinet temperatures. Different assets, same pattern: A quiet build-up that ends in a sudden-looking stop.
The assets are still running. The alarms are still quiet. And in most facilities, there is no system designed to see what is developing.
This is exactly why periodic inspection struggles to catch these patterns in time. A quarterly inspection, for instance, can leave roughly 89 days of degradation invisible between visits, which is plenty of time for a minor issue to become a stoppage. We’ll walk through a real example of exactly this gap in our upcoming webinar.
In a manual environment, one broken piece of equipment is a problem. In a conveyorized, automation-heavy environment, it's a chain reaction.
A single stoppage doesn't stay contained to one asset. It propagates upstream and downstream, idles labor that was synchronized around the line running, and puts the carrier cutoff window for the entire sort at risk. The interdependence that makes these systems efficient at full speed is the same interdependence that makes a single failure so expensive.
That changes what "reliability leader" actually means in these environments. It isn't the team running the newest equipment. It's the team that sees problems early enough to act before the cascade starts, the team for whom a developing fault is a work order, not an emergency.
Most conversations about condition monitoring start with questions around sensors, software, and maintenance systems. Those are the wrong starting questions. Detection isn’t a sensor or software choice—it’s a matter of when you find out something is wrong.
The question that actually determines your exposure is simpler: When do you find out?
Every operation sits somewhere on a maturity model for detection, from purely reactive (you find out when it stops), to alarm-based (you find out at or near the point of functional failure), to condition-aware (you're monitoring live changes across the asset), to continuous condition intelligence (you have consistent, ongoing visibility across your critical assets, prioritized by criticality). Where an operation sits in that model, more than any single piece of equipment, is what determines whether a $10,000-an-hour problem gets caught as a work order or lived through as a shift-ending event.
We call that model reliability maturity—the same framework behind the assessment.
That model is only useful once it's pointed at the right assets. Before any of the work around sensors, alerts, and monitoring tiers takes place, there's a more basic question to answer: Which assets actually matter enough to justify continuous visibility? Most detection gaps aren't a technology problem first. They're a prioritization problem. A criticality assessment—looking at single points of failure, throughput impact, safety exposure, and how hard the asset actually is to repair—is what tells you where to spend that attention before you spend a dollar on monitoring it.
Getting criticality and detection right doesn't just reduce downtime. In practice, it shows up in three places: safety (fewer people needing to work near live or elevated equipment, since continuous visibility replaces the routine check), uptime (early warning turns an unplanned stoppage into a repair scheduled on your own terms), and cost (labor and spare parts go toward the assets that actually need them, instead of a fixed schedule that treats every asset the same).
If you want to see where your own operation sits in that model, take our two-minute Reliability Maturity Assessment.
We're also covering this in a September 15 webinar hosted by Reliabilityweb, including the same self-assessment framework and a step-by-step roadmap for closing the gap once you know where you stand.
And if you want the full data behind the numbers in this post—the cost of late detection broken down by industry, and the four-tier maturity model referenced above—it's covered in our latest whitepaper.
Or, if you’re ready to see how multi-sensing monitoring can help your maintenance and reliability efforts, our team is eager to explain in a quick demo—schedule it here.
Detection latency is the time between when equipment degradation starts and when someone or something actually notices it. Every maintenance strategy has some amount of latency built in, whether or not it was designed that way on purpose.
Across industries, unplanned downtime costs roughly $125,000 per hour on average, according to ABB’s Value of Reliability Survey (2023). In distribution and sortation environments specifically, that figure runs $10,000 to $50,000 per hour, plus $15 to $50 per affected shipment once SLA cutoffs are missed.
Because not every asset carries the same risk if it fails. A criticality assessment, which means looking at single points of failure, throughput impact, safety exposure, and repair difficulty, determines which assets are worth continuous visibility first. Most detection gaps are a prioritization problem before they’re a technology problem.
Reactive means you find out when something stops. Alarm-based means you find out at or near the point of failure, when an alert fires. Condition-aware means you’re catching real early-warning signals because you have consistent visibility into how the asset is actually behaving.
A periodic inspection is only accurate at the moment it’s taken, because the asset’s condition can change again the moment the inspector walks away. A quarterly inspection, for example, can leave roughly 89 days of degradation invisible between visits, which is plenty of time for a minor issue to become a stoppage.
Start with the two-minute Reliability Maturity Assessment to see where your current approach sits and where the biggest gaps are, then use the Detection Latency Blueprint to understand the cost of those gaps in more detail.