Skip to content
Benchmarks
MetalsCycle-Time DriftFlag cycles inside a drift episode

Cycle-Time Drift Detection

Detect gradual cycle-time drift against validated standards from per-cycle timestamps. Higher F1 on drift episodes is better.

Metric

F1 (drift episodes)

Baseline

Cycle-Time Drift Detector

Submissions

28

Updated

2026-06-24

Leaderboard

RankSubmitterModelTypeScoreDate
1h.novakdriftnetResearch implementation0.9052026-06-24
2materealcycle-time-drift-detectorReference model0.8802026-06-18
3s.iyercusum-baselineBaseline model0.8122026-06-10
4d.moreauthreshold-baselineBaseline model0.7442026-06-02

Entries are reference and research submissions on the open split. Scores are illustrative and require plant-specific validation for industrial use.

Task definition

Evaluation metric

F1 (drift episodes)

Submission rules

  1. 1

    Train only on the provided synthetic training split; the test split is held out.

  2. 2

    Report F1 on labelled drift episodes with a fixed random seed.

  3. 3

    Submit the model card, training data type, and license with each entry.

Discussion

  • s.

    s.iyer

    ML researcher · 2d

    Question

    Are ensemble submissions allowed, or single-model only? Worth stating explicitly in the rules.

    7
  • ma

    matereal

    Maintainer · 2d

    Comment

    Ensembles are allowed if reproducible from the submitted artifacts. We'll clarify in the next rules revision.

    5
Open technical community

Contribute to open industrial AI

Browse open models, datasets, benchmarks, and process maps, or start a discussion. Customer plant data and private deployments stay confidential.