The methods behind Niobia AI.
Learn is the teach-first index for the analytical methods Niobia AI runs: root cause analysis, SPC, electrochemistry, DOE, materials characterization, testing workflows, and the platform docs that support them.
Platform docs7
SPC & Process Control8
RCA & 8D11
Electrochemical10
DOE & Predictive8
Materials Development8
Testing Methods5
Voice & Wearables2
Wafer Map AI2
Platform docs
Existing Niobia platform documentation, now grouped under Learn.
Getting started
Create a project, upload your first dataset, and generate your first plot.
Read more Platform docsHow to plot a graph
Upload data, choose any X/Y axis, switch plot types, and explore your curves interactively.
Read more Platform docsData requirements
Supported formats and the columns Niobia AI expects for ingestion and plotting.
Read more Platform docsCharge-discharge analysis
Analyze cycling behaviour, compare curves, and track the key battery metrics.
Read more Platform docsUsing the AI agent
Ask Niobia AI questions about your dataset, summarize trends, and spot anomalies faster.
Read more Platform docsFAQ & troubleshooting
Common issues with uploads, blank plots, column mapping, and better AI answers.
Read more Platform docsChangelog
Track recent updates to the platform docs and shipped product changes.
Read moreSPC & Process Control
See drift before it becomes scrap — capability, live charts, excursion alerts, multivariate.
Cp vs Cpk — why a centered process is not the same as a capable one
What Cp and Cpk actually mean, how to calculate them, why Cpk collapses when a process drifts off-center while Cp stays constant, the capability thresholds, and common mistakes.
Read more Method overviewEWMA
EWMA in Niobia AI's spc & process control workflows: Catches small, sustained shifts the raw chart misses.
Read more Method overviewExcursion alert
Excursion alert in Niobia AI's spc & process control workflows: A parameter trends out and the right engineer gets pinged.
Read more Method overviewGauge R&R
Gauge R&R in Niobia AI's spc & process control workflows: Can the gauge even see the signal you're charting?
Read more Method overviewHotelling T²
Hotelling T² in Niobia AI's spc & process control workflows: Two readings, each in spec — yet impossible together.
Read more Method overviewPp / Ppk
Pp / Ppk in Niobia AI's spc & process control workflows: Within-subgroup sigma vs the full long-term spread.
Read more Method overviewWestern Electric
Western Electric in Niobia AI's spc & process control workflows: Run rules on the sigma zones, not just the 3σ line.
Read more Method overviewX̄ / R chart
X̄ / R chart in Niobia AI's spc & process control workflows: Every sample against control limits, the moment it lands.
Read moreRCA & 8D
Troubleshoot defects in minutes, not months — guided 8D, fishbone, FMEA.
Corrective action
Corrective action in Niobia AI's rca & 8d workflows: Fix the verified root cause and watch the defect rate fall.
Read more Method overviewCross-batch RCA
Cross-batch RCA in Niobia AI's rca & 8d workflows: Pull SPC, materials and electrochemistry into one hypothesis.
Read more Method overviewFault tree
Fault tree in Niobia AI's rca & 8d workflows: How basic faults combine, through AND/OR gates, into the failure.
Read more Method overviewFMEA
FMEA in Niobia AI's rca & 8d workflows: Severity × Occurrence × Detection — prioritise by RPN.
Read more Method overviewInterim containment
Interim containment in Niobia AI's rca & 8d workflows: Screen out the bad parts now — before the cause is even known.
Read more Method overviewKnowledge capture
Knowledge capture in Niobia AI's rca & 8d workflows: Every closed investigation becomes searchable memory.
Read more GuidePareto analysis — finding the vital few defects worth fixing first
How to build and read a Pareto chart: ranked defect bars plus a cumulative line, the 80/20 principle, identifying the vital few, and the common mistakes that make a Pareto misleading.
Read more Method overviewPrevent recurrence
Prevent recurrence in Niobia AI's rca & 8d workflows: Lock it in so the same failure can't come back.
Read more GuideThe 5 Whys — drilling from symptom to systemic cause
How to use the 5 Whys to drill from a symptom to a systemic root cause, when five is the wrong number, how to avoid jumping to blame, and how it pairs with the fishbone and 8D.
Read more GuideThe 8D problem-solving process, without the box-ticking
A practical walkthrough of the 8D method: all nine disciplines from D0 to D8, why containment is not correction is not prevention, the common failure modes, and how to run 8D faster.
Read more GuideThe fishbone diagram and the 6Ms — structured brainstorming for root cause
How to build a fishbone (Ishikawa) cause-and-effect diagram using the 6M categories — Man, Machine, Method, Material, Measurement, Mother Nature — to structure root-cause brainstorming.
Read moreElectrochemical
From potentiostat and cycler files to cell-level diagnosis in one pass.
Capacity fade
Capacity fade in Niobia AI's electrochemical workflows: Capacity retention against cycle number, with the 80% end-of-life line.
Read more GuideCharge/discharge curves and CCCV — reading a cell's voltage profile
How to read galvanostatic charge/discharge curves: voltage plateaus and what they mean, the CCCV protocol and its constant-voltage taper, coulombic efficiency, and how capacity fade shows up cycle over cycle.
Read more Method overviewDegradation decomposition
Degradation decomposition in Niobia AI's electrochemical workflows: Total fade, split into the mechanisms that caused it.
Read more Method overviewDiffusion reconciliation
Diffusion reconciliation in Niobia AI's electrochemical workflows: Three methods, one diffusion coefficient.
Read more GuideGITT and PITT — getting a diffusion coefficient out of a battery
How the galvanostatic and potentiostatic intermittent titration techniques (GITT/PITT) work: the pulse-relaxation staircase, why the relaxation transient carries the diffusion coefficient, and how D varies with state of charge.
Read more GuideHow to read a cyclic voltammogram — and diagnose a cell from it
A practical guide to cyclic voltammetry: the duck-shaped I–E curve, peak separation and reversibility, the square-root scan-rate law, common pitfalls, and how to diagnose electrode kinetics from a CV.
Read more GuideHow to read a Nyquist plot — and see a cell age in real time
A practical guide to the EIS Nyquist plot: the high-frequency intercept (Rs), the charge-transfer semicircle (Rct), the 45° Warburg tail, the Randles circuit, and how a growing semicircle reveals cell aging.
Read more Method overviewLinear sweep
Linear sweep in Niobia AI's electrochemical workflows: One-way sweep to a diffusion-limited plateau.
Read more Method overviewNucleation & growth
Nucleation & growth in Niobia AI's electrochemical workflows: Nuclei appear, grow, and merge while the current traces a hump.
Read more GuideThe Bode plot and the Randles circuit — the rest of an EIS analysis
How the Bode plot and the Randles equivalent circuit complete an EIS analysis: reading magnitude and phase versus frequency, locating the RC time constant, and mapping each circuit element to a feature of the impedance spectrum.
Read moreDOE & Predictive
Design fewer experiments, learn more from each — designs, RSM, active learning.
Bayesian optimization
Bayesian optimization in Niobia AI's doe & predictive workflows: Niobia recommends the next experiment to run.
Read more Method overviewBox-Behnken
Box-Behnken in Niobia AI's doe & predictive workflows: Edge midpoints — never the extreme corners.
Read more GuideCentral composite design — fitting curvature without running everything
How a central composite design works: the factorial cube, axial (star) points, and center points, why they let you fit curvature with a quadratic model, rotatability and alpha, and how CCD compares to Box-Behnken.
Read more Method overviewFull factorial
Full factorial in Niobia AI's doe & predictive workflows: Every factor combination — full information.
Read more GuideGaussian processes — a model that knows where it's ignorant
How Gaussian process regression (kriging) models a response with calibrated uncertainty: the mean prediction, the confidence band that pinches at data and balloons between, the kernel, and why it powers active learning and Bayesian optimization.
Read more Method overviewLatin hypercube
Latin hypercube in Niobia AI's doe & predictive workflows: Space-filling: one sample per row and column.
Read more Method overviewPrediction profiler
Prediction profiler in Niobia AI's doe & predictive workflows: Drag a factor, watch the prediction move everywhere.
Read more GuideResponse surface methodology — turning runs into a process window
How response surface methodology turns experimental runs into a fitted surface and contour map: reading the surface, the quadratic model, finding stationary points (peak, valley, saddle, ridge), and steepest ascent.
Read moreMaterials Development
Turn XRD, XPS, SEM and Raman into decisions, not slide decks.
AFM
AFM in Niobia AI's materials development workflows: Nanoscale topography, quantified.
Read more Method overviewFTIR
FTIR in Niobia AI's materials development workflows: Absorption bands fingerprint the functional groups.
Read more GuideHow to read an XRD pattern — from Bragg's law to phase ID
How X-ray diffraction works: Bragg's law (nλ = 2d·sinθ), why peaks appear at specific 2θ angles, how peak positions index crystal planes for phase identification, and what peak width and intensity mean.
Read more Method overviewParticle size
Particle size in Niobia AI's materials development workflows: The spread often matters more than the mean.
Read more Method overviewRaman
Raman in Niobia AI's materials development workflows: Each peak is a specific molecular vibration mode.
Read more Method overviewRietveld
Rietveld in Niobia AI's materials development workflows: Fit the whole pattern; watch the difference flatten.
Read more Method overviewSEM / TEM
SEM / TEM in Niobia AI's materials development workflows: Turn a micrograph into particle statistics automatically.
Read more Method overviewXPS
XPS in Niobia AI's materials development workflows: Surface chemistry and oxidation state, not just elements.
Read moreTesting Methods
From raw instrument data to a publishable report in minutes.
Auto-plot suite
Auto-plot suite in Niobia AI's testing methods workflows: A full plot suite, chosen and drawn automatically.
Read more Method overviewAuto-report
Auto-report in Niobia AI's testing methods workflows: Methods, plots and conclusions, on brand.
Read more Method overviewIngestion
Ingestion in Niobia AI's testing methods workflows: Drop in instrument files; Niobia parses and checks them.
Read more Method overviewSpec compliance
Spec compliance in Niobia AI's testing methods workflows: Checked against internal and customer specs, instantly.
Read more Method overviewStress–strain
Stress–strain in Niobia AI's testing methods workflows: The canonical curve — yield, UTS, fracture.
Read moreHardware
Connect Niobia to your lines and instruments.
Instrument connectors
Instrument connectors in Niobia AI's hardware workflows: potentiostats · cyclers · XRD
Read more Method overviewNiobia hardware box
Niobia hardware box in Niobia AI's hardware workflows: on-prem bridge
Read more Method overviewPLC / OPC-UA / MQTT
PLC / OPC-UA / MQTT in Niobia AI's hardware workflows: connect lines
Read moreVoice & Wearables
Hands-free troubleshooting on the line.
Wafer Map AI
Wafer-map intelligence for defect localization and excursion diagnosis.