Practical explainers of the methods behind the software, plus product updates from the Reliafy team. Looking for how to do something in the app? See the guides.
Every serious reliability package assumes you'll take the course — the licence is the small half of what it costs to make an engineer productive. Reliafy ships the expertise inside the tool instead, as an assistant that helps you drive and an agent that does the analysis with you, then asks permission before it saves anything.
Read more →Every mature RCM programme has the same artifact — a thorough, expensive study that was true the day it was written and has been quietly diverging from reality ever since. The fix isn't more discipline. It's making the worksheet check itself.
Read more →A fitted life model tells you how one component behaves. The question your planner actually asks is about the fleet — "how many of these fail in the next twelve months?" Here's the right way to answer it, and the new Fleet section in Reliafy that does it for you.
Read more →Most equipment doesn't fail out of nowhere — it wears, measurably. Degradation analysis turns those measurements into a life model for the population and a failure-time estimate for every individual asset you're running today. Here's how it works, and how to do it in Reliafy.
Read more →The full reliability-engineering toolkit — modelling, RBDs, and maintenance strategy — is now AGPL-licensed and self-hostable with one command. Here's what changed, why we did it, and exactly how the open-source and cloud versions differ.
Read more →One place to turn failure data into fitted life models, reliability block diagrams, and cost-optimal maintenance decisions.
Read more →What Weibull analysis is, what the shape and scale parameters actually tell you, how fitting works (with a worked example), how censored data changes everything, and how the results turn into maintenance decisions.
Read more →ALT tests units at elevated stress to fail them faster, then extrapolates back to normal use with a life-stress model. Here are the Arrhenius, inverse-power and Eyring relationships, the acceleration factor, and a worked temperature example.
Read more →B10 life is the time by which 10% of a population has failed. Here's the exact Weibull formula, a worked calculation, how B10 relates to bearing L10, and why it beats MTBF for setting intervals.
Read more →Most reliability datasets are dominated by units that haven't failed. Ignore those suspensions and your life estimates come out 2× wrong or worse. Here's what censoring is, with a worked example of the damage.
Read more →Redundancy only helps if the redundant units fail independently. Common-cause failures break that assumption. Here's the beta-factor model, a worked example of how much it erodes redundancy, and when to use MGL instead.
Read more →A complete, honest walkthrough of Weibull analysis in Excel using median rank regression — worked example included — and the exact point where Excel stops being the right tool.
Read more →In a load-sharing system, surviving units pick up the load of failed ones and fail faster as a result — so parallel-redundancy math overstates reliability. Here's the load-life mechanism, a worked example, and how it differs from standby.
Read more →MTBF is for repairable systems, MTTF for non-repairable — that's the textbook line. Here's the full story: exact definitions, how to compute each from real data, and the ways both numbers routinely deceive.
Read more →A complete reliability centred maintenance example — functions, functional failures, failure modes, consequences, and task selection for a cooling-water pump — with the worksheet filled in and the reasoning shown.
Read more →Availability is the fraction of time a repairable system is up. Here's the inherent-availability formula from MTBF and MTTR, how it composes through a reliability block diagram, why redundancy helps, and how to read per-component downtime.
Read more →