This is the hub for a nine-part series on disciplined fault diagnosis — one long argument told in order, from a single vehicle’s pile of untrustworthy clues all the way to a fleet-wide remediation campaign and back around to engineering the fault out for good.

The visual map

Open the series map →

The nine parts

I · The core method

  1. Root-Cause Analysis Under Uncertain, Heterogeneous Evidence — grade, anchor, and fuse messy, unequal-trust clues into one defensible cause.
  2. When Is the Evidence Enough? — the sufficiency gate, value of information, the stopping test, and handling conflict.
  3. Harder Cases in Diagnosis — multiple faults, intermittency, and lying sensors.

II · At fleet scale

  1. Fleet-Scale Root-Cause Analysis — de-mix the mixture, let the cohort boundary name the cause, decide a campaign.
  2. Fleet RCA, Made Rigorous — significance, surveillance, forecasting, and closing the loop.

III · Proof, foundations & prevention

  1. Proving Cause with Experiments — DAGs, staged rollouts as randomized experiments, difference-in-differences, instrumental variables.
  2. Diagnosis on the Vehicle — real-time detection at the edge, feeding the fleet pipeline.
  3. The Data Plumbing Behind Fleet Diagnosis — joining, aligning, and cleaning the data that makes it all possible.
  4. From Root Cause to No Cause — prevention and the corrective-action loop.

IV · Build the system

  1. From Method to Product — the architecture of an AI-powered, multi-agent fleet-diagnosis tool: data, detection, agents, tools, memory, and the human in the loop.

Reference

The Diagnosis & Prognosis Glossary — every term used across the series, from cohort and DTC to dominant stressor, confounder, residual, and RUL, defined in plain English and grouped into thirteen themes. Keep it open in a second tab.

Where to start

New to the topic? Read 1 → 2 → 3 for the reasoning discipline, then 4 → 5 for fleets. Already do this work and want the sharp edges? Jump to 5 (rigor), 6 (proof), or 9 (prevention). Building the system rather than doing the analysis? Start with 8 (data) and 7 (edge). New to the vocabulary? Start with the glossary.

← Back to Autonomy