Reliability Engineer — turning failure data into uptime and safety.
20+ years across three of the most demanding industries there are: aerospace (Embraer, Azul) → mining (Vale) → subsea oil & gas NPD (OneSubsea / SLB).
I work at the intersection of reliability engineering, data science, and functional safety: life-data and degradation analysis, failure and root-cause investigation, and designing critical systems so that the safe state is the default state.
🔬 FLARE26 — an auditor that knows when not to answer
A neuro-symbolic, glass-box system that compares documents and abstains when the requested information is genuinely absent, or of the wrong type or scope — then decides consensus vs. divergence with a deterministic judge.
It targets the most expensive error in multi-document audit: a false-positive divergence, a disagreement the system invents by hallucinating an answer the document never gave.
Reported transparently, negative results and limitations included.
Python · MIT · live demo (no API key) · DOI: 10.5281/zenodo.20881699
📐 relengy — reliability engineering in Python
Life data analysis, reliability growth, accelerated testing, RBD, FMEA and fault trees, built against Abernethy's The New Weibull Handbook.
What separates it from the alternatives is that the method decisions are written down where the code makes them. Where the standard authorities disagree, the library says which side it took, and why.
Python · MIT
🌡️ ten-degree-rule — the shortcut that hides an activation energy
The electrolytic-capacitor rule that life doubles for every 10 °C of cooling is taught as mechanism-free. Solved backwards, it is an Arrhenius model, and the activation energy it implies varies 77% (0.549 to 0.971 eV) depending on where on the temperature scale you apply it.
Over the usual 105 °C → 40 °C extrapolation it embeds Ea = 0.707 eV, within 4% of the
only published value for that exact capacitor technology. So the rule is right by
coincidence, not by construction: nothing in it stops you carrying it to a neighbouring
technology, where the same error runs from 2.40× optimistic to 0.13× conservative.
Three competing manufacturers confirm the bound in their own technical notes. Closed-form, reproducible with the standard library, and a scope error in an earlier version of the analysis is corrected in the open rather than quietly.
Python · MIT
🧮 weibayes-zero-failures — what you can claim when nothing has failed
60 units, 79,371 accumulated hours, zero failures. A two-parameter Weibull has nothing
to fit, so the answer rests on a decision (the assumed shape β) as much as on data. This
study quantifies exactly how much.
It publishes the sufficient statistics instead of the individual lifetimes, which keeps
the underlying data closed while leaving every calculation checkable: verify.py
reproduces every published number with no access to the raw lifetimes and no dependencies.
Python · MIT
Reliability modelling for a PhD thesis on shell-and-tube heat exchangers: OREDA-2015 reports only aggregate exposure time and failure counts, never unit-level histories. We reconstruct statistically consistent time-to-failure sequences via a Poisson renewal process, select a lifetime model by likelihood-based criteria, and derive RCM maintenance intervals.
The load-bearing result is negative. We prove that the fitted Weibull shape, and the Weibull-over-Exponential model choice itself, are artefacts of the reconstruction geometry rather than evidence of wear-out, so the shape is carried as an explicit sensitivity assumption instead of reported as a finding.
Collaborator and co-author. Manuscript in preparation.
Reliability — RAM analysis · FMEA/FMECA · RCA · FTA · RCM · life data & degradation analysis · accelerated life testing · optimum replacement interval
Safety — functional safety · HAZOP · design for reliability
Data — Python (pandas, numpy, scipy, reliability, matplotlib) · R (WeibullR) · SQL
Commercial — ReliaSoft Weibull++ · BlockSim · ALTA
Standards — ISO 14224 · MIL-STD-1629A
Certified Reliability Professional (CRP) · Certified Functional Safety Professional · Reliability Consultant
Specializations: Reliability Engineering (UTFPR) · Big Data & Data Science (Mackenzie) · Aviation Safety (ITA) · Engineering Innovation & Entrepreneurship (UNESP)
I write about reliability, failure, and the logic behind the data on LinkedIn.