Key takeaways
Short answer: Availability and reliability are related but not the same. Availability is an outcome — what fraction of planned time the asset was up. Reliability is a behaviour — how long it runs between failures. A machine that fails often but is fixed in minutes can post high availability while being deeply unreliable, and that distinction matters because the two problems have different fixes. For how availability rolls into the bigger picture, see OEE for manufacturing.
Availability answers a blunt question: of the time we planned to run this asset, how much was it actually up and capable of producing? It is uptime divided by planned production time, usually expressed as a percentage. It is an aggregate — it does not care whether the lost time came from one long stoppage or fifty short ones, only the total. That makes it easy to calculate and easy to communicate, and it is exactly the figure that feeds the availability term in OEE. But because it only reports the total, availability alone can mask very different underlying behaviours.
Reliability is about behaviour over time: how long does the asset run before it fails, and how predictable is that? The standard measure is MTBF — mean time between failures — the average run length between breakdowns. A reliable asset has long, consistent intervals between failures; an unreliable one fails often, erratically, or both. Reliability says nothing directly about how fast you recover; that is repairability, measured by MTTR (mean time to repair). Two assets can share the same availability while one fails monthly and the other fails daily — and they are not the same risk at all.
Here is the trap. Imagine two machines, both at 95% availability. Machine A failed once and was down for an afternoon. Machine B failed thirty times, each for a few minutes, and was nursed back each time by a quick reset. Same availability, completely different health. Machine B is unreliable: it is eating operator attention, signalling a developing fault, and one of those resets will eventually not work — turning a string of micro-stops into a major breakdown. If you only watch availability, Machine B looks fine until the day it isn't. Reliability metrics are what surface the difference before it bites.
The two metrics behind the story are simple. MTBF = total run time ÷ number of failures; a higher MTBF means a more reliable asset. MTTR = total repair time ÷ number of failures; a lower MTTR means faster recovery. Availability is then roughly MTBF ÷ (MTBF + MTTR). That formula is why fast repairs can prop up availability even as reliability rots: shrink MTTR enough and availability stays high while MTBF quietly falls. Worked example — an asset with MTBF of 40 hours and MTTR of 2 hours runs near 95% availability; if MTBF drops to 8 hours but crews cut MTTR to 25 minutes, availability still reads near 95% while the asset now fails five times as often. The number lies; the trend in MTBF tells the truth.
The diagnosis drives the fix. If availability is low and MTBF is also low, you have a reliability problem — chase root cause: lubrication, alignment, wear, design. If availability is low but MTBF is healthy and MTTR is high, you have a repairability problem — chase spares, access, procedures, and skills, not the asset itself. Pursuing the wrong one wastes money: buying a more reliable part will not help a machine that is reliable but slow to repair, and faster repairs will not save a machine that simply keeps breaking. Always read the two together.
Availability is one of the three factors in OEE, alongside performance and quality. Reliability is the behaviour underneath it. Crucially, low reliability does not only hurt the availability term — the frequent micro-stops that mark an unreliable asset also drag down performance, because the line never settles into its rated speed. That is why an unreliable machine often shows a double OEE hit. Tracking failures, not just downtime totals, lets you connect a sagging OEE to its real cause and target the losses properly.
Fabrico tracks both sides. It measures availability against planned time for live OEE, and it records every stop and failure — so MTBF and MTTR fall out of the same data, and a high-availability asset that is quietly becoming unreliable is flagged before it fails hard. Seeing failure frequency next to availability is what stops a clean-looking number from hiding a developing breakdown. Book a demo to see availability and reliability side by side on your equipment.
Availability is the share of planned time an asset is up and ready to run. Reliability is how long it runs before failing, usually measured as MTBF. High availability can coexist with poor reliability if failures are frequent but quickly repaired.
Availability is uptime divided by planned production time, expressed as a percentage. In terms of failure metrics it is approximately MTBF ÷ (MTBF + MTTR).
MTBF, mean time between failures, is total run time divided by the number of failures. It is the standard measure of reliability — a higher MTBF means longer, more consistent runs between breakdowns.
Yes. If a machine fails frequently but is repaired very quickly, its total downtime stays small and availability looks high, even though it is unreliable. Tracking failure counts and MTBF reveals the difference.
Availability is one of the three OEE factors. Reliability is the behaviour driving it, and because frequent micro-stops also cut into the performance factor, low reliability often lowers OEE twice over.
Login
Blog
