Anomaly Detection on Sensor, SCADA and Historian Data

PCA, MSET, isolation forest, autoencoders and small pretrained time-series models; AWS Lookout for Equipment ends 7 Oct 2026, replaced by SiteWise native anomaly detection.

time series

5named models

63candidates

1deployments

1in Indonesia

Named Models

MOMENT (MOMENT-1-large)Carnegie Mellon UniversityReleasedFeb 2024Runs onsingle datacentre GPUClasstime series

General-purpose open time-series model for forecasting, classification and anomaly detection; not submitted to GIFT-Eval. Last PyPI momentfm 0.1.4 on 19 Mar 2025.Commercial use: allowed; Jakarta hosting: unknown; Indonesian support: not stated Conv-AEnot in the notesReleasednot in the notesRuns onnot in the notesClasstime seriesConvolutional autoencoder anomaly detector, top of the SKAB outlier leaderboard.Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not stated

Granite TSPulseIBMReleasedMay 2025Runs onCPU serverClasstime series

IBM tiny time-series model for anomaly detection and more (paper 19 May 2025, ICLR 2026). Claims #1 on TSB-AD univariate and multivariate (+20%) with no scores on the card. Maximo integration unverified.Commercial use: allowed; Jakarta hosting: unknown; Indonesian support: not stated MSET (Multivariate State Estimation Technique)not in the notesReleasednot in the notesRuns onnot in the notesClasstime seriesSimilarity-based anomaly detection method; widely associated with GE SmartSignal (unverified).Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not stated

PaAnoSungkyunkwan UniversityReleasednot in the notesRuns onnot in the notesClasstime series

Time-series anomaly detector ranked 1 on TSB-AD (VUS-PR 0.53 U, 0.43 M) in its ICLR 2026 paper.Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not stated

Where this comes from

Models the research names for this use case. Research, with sources on each model record.

Who Runs It

production, 1

PT Pertamina Hulu Rokan (PHR)Country: IndonesiaDICE (Digital and Innovation Center): AI-supported big data centre monitoring thousands of wells in near real time on 66 screens at Rumbai, Pekanbaru (model not stated)Vendor: not statedproductionDate17 Oct 2025Production decline at WK Rokan held from 11 percent a year to 0 percent (company claim)Evidence: company statementConfidence: lowSource: kabarbursa.com, en.wikipedia.org

Where this comes from

Documented uses by named organisations. The evidence column says who made the claim.

Candidate Models

TSB-AD

A general benchmark for this model class, not a test on this use case. Time-series anomaly detection, 1,070 series from 40 datasets (VUS-PR); NeurIPS 2024; open to community submissions from 1 Apr 2026. Higher is better.

PaAnoSungkyunkwan University vendor-reported0.53 (Sub)-PCA independent0.42

MOMENT (MOMENT-1-large)Carnegie Mellon University independent0.39 DeepAnT independent0.31 OmniAnomaly maintainer-run0.31 PCA independent0.31

00.51

NAB (Numenta Anomaly Benchmark)

A general benchmark for this model class, not a test on this use case. Streaming univariate anomaly detection; legacy, little maintenance. NAB score. Higher is better.

ARTime maintainer-run74.9 Numenta HTM maintainer-run70.5 CAD OSE maintainer-run69.9

050100

TimesFM 3.0Google, time seriesReleasedAug 2026open weightshas scoreslast 18 months

Toto 2.0 1BDatadog, time seriesReleasedJun 2026open weightshas scoreslast 18 months

Toto 2.0 2.5BDatadog, time seriesReleasedJun 2026open weightshas scoreslast 18 months

Toto 2.0 313MDatadog, time seriesReleasedJun 2026open weightshas scoreslast 18 months

Granite PatchTST-FM r1IBM, Rensselaer Polytechnic Institute, time seriesReleasedMar 2026open weightshas scoreslast 18 months

Moirai 2.0 smallSalesforce, time seriesReleasedNov 2025open weightshas scoreslast 18 months

Chronos-2Amazon, time seriesReleasedOct 2025open weightshas scoreslast 18 months

FlowState r1.1IBM, time seriesReleasedSep 2025open weightshas scoreslast 18 months

TimesFM 2.5Google, time seriesReleasedSep 2025open weightshas scoreslast 18 months

TiRexNX-AI, time seriesReleasedMay 2025open weightshas scoreslast 18 months

Showing 10 of 63. All models of this class

Where this comes from

Derived, not researched: models in the directory whose class matches a class this use case uses, excluding family records and models shut down or deprecated. Ranked by how many fit facts they meet, then by release date. Each fit fact is read from the model record.

Use Case

Industry: industrial and asset-intensive
Model classes: time series
Benchmarks or datasets: TSB-AD SKAB NAB (Numenta Anomaly Benchmark)
Regulatory notes: Surat Edaran Menkominfo 9/2023 Etika Kecerdasan Artifisial, Kementerian Komunikasi dan Informatika (now Kementerian Komunikasi dan Digital)Anomaly detection on sensor, SCADA and historian data is an AI-based service run by an electronic system operator, and the circular asks such operators to build and use AI on the values of inclusivity, transparency, humanity and security.

Where this comes from

From the industry tracks T08 to T10 of the research notes.

R&D Topics

Links to R&D topics are awaiting review.

Where this comes from

Which R&D topics would most change this use case. Drafted by Claude from the topic definitions and the use case; proposed until approved.

Incidents

Nuclear False AlarmRussia, 1983-09-26, safetyAn alert of five incoming intercontinental ballistic missiles was properly identified as a false-positive by the Soviet Union operator Stanislov Petrov.

Where this comes from

Documented failures, harms or misuse in this kind of use, Indonesia first, then the region, then well-known cases elsewhere (AI Incident Database, OECD AI incidents monitor, reputable news). Not a claim about any deployment listed here.

Opportunities

Opportunities are awaiting review.

Where this comes from

Gaps found by rule in what the directory holds for this use case. Each is a proposal for review.

Links

IndustryIndustrial30 use cases, 54 deployments

Use CaseAnomaly Detection on Sensor, SCADA and Historian DataIndustrial

Models

Granite TSPulseIBM, time series

MOMENT (MOMENT-1-large)Carnegie Mellon University, time series

PaAnoSungkyunkwan University, time series Conv-AEtime series+1 more

Who Runs It

PT Pertamina Hulu Rokan (PHR)In production, 17 Oct 2025

Graph

Models 5

Granite TSPulseIBM, time series

MOMENT (MOMENT-1-large)Carnegie Mellon University, time series

PaAnoSungkyunkwan University, time series Conv-AEtime series MSET (Multivariate State Estimation Technique)time series

Deployments 1

PT Pertamina Hulu Rokan (PHR)DICE (Digital and Innovation Center): AI-supported big data centre monitoring thousands of wells in near real time on 66 screens at Rumbai, Pekanbaru (model not stated), Indonesia

Sources

ibm-granite/granite-timeseries-tspulse-r1 , Hugging Face huggingface.co https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1
AutonLab/MOMENT-1-large , Hugging Face huggingface.co https://huggingface.co/AutonLab/MOMENT-1-large
arxiv.org/pdf/2602.01359 https://arxiv.org/pdf/2602.01359
GitHub - waico/SKAB: SKAB - Skoltech Anomaly Benchmark. Time-series data for evaluating Anomaly Detection algorithms. github.com https://github.com/waico/SKAB
Predictive Maintenance - Amazon Lookout for Equipment - AWS aws.amazon.com https://aws.amazon.com/lookout-for-equipment/

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Anomaly Detection on Sensor, SCADA and Historian Data", https://aurion.semai.id/usecases/ind-sensor-anomaly-detection/, records of October 2026.