Skip to content
Contact UsStart Building

Transport & Fleet Maintenance

Truck fault diagnosis

Identify recorded air-pressure-system faults from truck measurements to guide diagnostic review.
AI-generated representative illustration of a heavy-truck pneumatic air reservoir with physically credible valves, fittings and hoses.
AI-generated task illustration. Representative scene; not a customer site or benchmark observation.
From known examples to a prediction

Available inputs

  • Anonymized operating measurements
  • Histogram-style sensor summaries

Known examples

76,000 recorded examples

Prediction target

The recorded air-pressure-system failure category
0.24 pphigher ROC AUC

vs. XGBoost (tuned ensemble), the strongest conventional result in this benchmark.

Truck air-pressure system failuresROC AUC · Higher is better

76,000 examples · 170 inputs · 9 evaluation splits

01
  1. Causilo
    0.9949
  2. XGBoost
    0.9925
  3. CatBoost
    0.9924
  4. LightGBM
    0.9923
  5. Extra trees
    0.9911
  6. Random forest
    0.9904
  7. K-nearest neighbors
    0.9897
  8. Logistic regression
    0.9896

The task in practice

Anonymized truck measurements and sensor summaries are labeled by whether a recorded fault involved the air-pressure system. The task distinguishes fault categories to evaluate a diagnostic review signal.

  1. Connect measurements with failure labels. Use historical vehicle profiles and their recorded failure categories.
  2. Describe another measurement profile. Supply the same numerical summaries in the curated schema.
  3. Assess a diagnostic review signal. Compare missed failures and false alerts before evaluating a maintenance workflow.

Scope. The task classifies recorded failure categories; it does not establish lead time before a failure occurs.

Dataset, sources and evaluation

Truck air-pressure system failures dataset. Original source and curated OpenML task linked below. Original dataset source.

The TabArena-curated OpenML version (46908) supplies the dataset and recorded target, AirPressureSystemFailure. Comparison scores come from the retained results_per_split_official.csv evaluation snapshot. These pages describe the task and compare the recorded means; source records are not republished here.

ROC AUC measures how well the scores rank the recorded positive outcomes above negative ones. A higher score is better. The change is reported in percentage points, rather than relative error reduction.

Causilo uses its default configuration. Each conventional family uses its strongest reported mean across default, tuned and tuned-ensemble settings. The chart compares reported means, without a claim of statistical significance, customer deployment or operational savings.

ROC AUC difference in percentage points = (Causilo mean − strongest conventional mean) × 100. ROC AUC is a ranking metric, not a percentage of correct predictions.

How to read these comparisons · All configurations and original values.