Predictive Maintenance MCP Server

Give your AI assistant evidence-based vibration diagnostics — machinery fault detection, ISO-cited severity, and diagnostic reports built to support and accelerate expert decision-making.

An open-source MCP server that turns LLMs into condition monitoring assistants for reliability engineers. Its core design rule: the server refuses to guess. No diagnosis is ever inferred from filenames or statistical parameters alone — a fault indication requires matching spectral evidence. Every severity claim cites ISO 20816-3, and the evaluative wording in reports is authored by the server, not improvised by the model. The AI orchestrates the analysis and presents the evidence — detected fault frequencies, matched fault patterns, severity zones — while the final judgment stays with the engineer. Also available as a Claude Code plugin with 8 diagnostic skills.

See It in Action

Choose Your Path

Quick Start

Get running in ~3 minutes. On Windows, one script wires everything into Claude Desktop — it installs the venv, pre-compiles dependencies, and writes claude_desktop_config.json for you (OneDrive / cloud-sync paths included):

git clone https://github.com/LGDiMaggio/predictive-maintenance-mcp.git
cd predictive-maintenance-mcp
.\setup_claude.ps1

Restart Claude Desktop, then try:

"Load real_train/OuterRaceFault_1.csv and check if the bearing is healthy."

Install the package:

pip install predictive-maintenance-mcp

Find the full path to uvx (which uvx on macOS/Linux, where uvx on Windows), then add to your client config — ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "predictive-maintenance": {
      "command": "/full/path/to/uvx",
      "args": ["predictive-maintenance-mcp"],
      "env": { "UV_LINK_MODE": "copy" }
    }
  }
}

Why the full path? Claude Desktop launches servers with a minimal PATH that often omits user-local tool directories (e.g. ~/.local/bin). Using the full path to uvx avoids a silent "command not found" failure. On Windows the typical path is C:\Users\<you>\.local\bin\uvx.exe.

More options: install from source · VS Code setup · Docker / HTTPS deployment · use with local LLMs (Ollama)

Benchmark

A blind, reproducible diagnostic-accuracy benchmark on the public CWRU Bearing Data Center dataset (12 kHz drive-end subset: 60 fault records + 4 normal baselines). Fault labels never reach the system under test — signals enter under opaque ids, a separate scorer is the only label reader, and blindness, checksum integrity, and determinism are enforced by CI-run guard tests, not prose. Results are stratified by the per-record diagnosability grades of the Smith & Randall (2015) reference study, so records that study found undiagnosable by any classical method are reported separately instead of inflating or deflating the headline.

On records the reference study grades clearly diagnosable (Y1+Y2, <!-- slot: headline.n_records -->44<!-- /slot --> records): characteristic fault frequency detected on <!-- slot: headline.frequency_detection.hits -->44<!-- /slot -->/<!-- slot: headline.frequency_detection.total -->44<!-- /slot -->, correct fault ranked first on <!-- slot: headline.classification.correct -->34<!-- /slot -->/<!-- slot: headline.classification.total -->44<!-- /slot --> (<!-- slot: headline.classification.rate pct1 -->77.3<!-- /slot -->%), and <!-- slot: strata.Y1.classification.correct -->9<!-- /slot -->/<!-- slot: strata.Y1.classification.total -->9<!-- /slot --> on the textbook-signature (Y1) stratum. On the <!-- slot: strata.ungraded.false_positives.total_normal -->4<!-- /slot --> healthy baselines, <!-- slot: strata.ungraded.false_positives.records_with_any -->2<!-- /slot --> records raised a false indication under the same criterion.

The numbers above are read from the committed, re-runnable artifact (results.json) and drift-guarded by CI: every value is bound to its key in the artifact, and a mismatch fails the build. Methodology, blind protocol, and honest-benchmarking notes: docs/benchmark-methodology.md. Reproduce with:

python -m benchmarks.cwru all

What Can It Do?

Point the AI at a vibration signal → get the evidence behind the fault — detected frequencies, matched fault patterns, ISO-cited severity — to support your call.

The AI doesn't guess — it calls 37 specialized MCP endpoints (34 tools + 3 prompts) running locally on your machine. Every signal is referenced by a single signal_id handle from load to report. Your data never leaves your infrastructure.

Full endpoint reference, grouped by category: Tool Catalog.

Claude Code Plugin

The project includes a plugin for Claude Code with domain-specific skills that activate automatically during conversation.

/plugin marketplace add LGDiMaggio/predictive-maintenance-mcp
/plugin install predictive-maintenance@predictive-maintenance-marketplace

The plugin adds 8 skills that activate automatically based on context (bearing-diagnosis, gear-diagnosis, quick-screening, report-generation, anomaly-detection, signal-management, documentation-search, prognostics), 2 agents that run multi-step diagnostic workflows end-to-end and hand you the evidence (diagnostic-pipeline, signal-explorer), and 3 commands for quick entry points (/pm-diagnose, /pm-screen, /pm-report).

Full skill, agent, and command reference: Plugin README.

Reports

All analysis tools generate interactive HTML reports you can open in any browser — pan, zoom, hover for details. Also supports structured Word (.docx) exports.

Sample Data Included

The project ships with 20 real bearing vibration signals from production machinery tests — ready to use out of the box: a training set (2 healthy baselines + 12 fault signals, inner and outer race) and a test set (1 healthy baseline + 5 fault signals).

Try: "Load real_train/OuterRaceFault_1.csv and diagnose the bearing fault."

Full dataset documentation: data/README.md

Architecture

          YOU (natural language)
               │
               v
     LLM (Claude, GPT, Ollama...)
     understands intent, selects tools
               │
               v  ── Model Context Protocol ──
    ┌──────────────────────────────┐
    │    Predictive Maintenance    │
    │         MCP Server           │
    │                              │
    │  Signal Analysis    Reports  │
    │  Fault Detection    ML       │
    │  Severity Rating    RAG Docs │
    └──────────────────────────────┘
               │
               v
       YOUR DATA (stays local)
    signals · manuals · models

The codebase follows a modular architecture organized around the ISO 13374 Six-Block Diagnostic standard — signal acquisition, processing, diagnostics, prognostics, and decision support as separate sub-packages. Standards implemented: ISO 13374, ISO 20816-3, MIMOSA OSA-CBM. Module-level detail: Architecture guide.

Key design choices:

• Privacy-first — raw vibration data never leaves your machine; only computed results flow to the LLM
• LLM-agnostic — works with Claude, ChatGPT, Microsoft Copilot Studio, or any MCP-compatible client. Use Ollama for fully air-gapped deployments
• Modular — use only the tools you need, extend with your own

Documentation

Testing

85%+ test coverage, enforced as a CI minimum, across Windows, macOS, and Linux (Python 3.11 & 3.12) — the current measured figure is on the codecov badge above.

pytest                                  # run all tests
pytest --cov=src --cov-report=html      # with coverage report

20+ test files covering signal analysis, fault detection, severity assessment, ML models, report generation, RAG search, and real bearing fault data validation.

Roadmap

• 37 MCP endpoints (34 tools, 3 prompts) with modular architecture and a single signal_id handle
• Claude Code plugin (8 skills, 2 agents, 3 commands)
• 85%+ test coverage enforced in CI, CI/CD on 3 platforms
• Docker + SSE/HTTP transport for enterprise deployment
• Semantic document search (FAISS + TF-IDF)
• Blind, reproducible diagnostic benchmark on the CWRU dataset (extensible to Paderborn)
• Customizable severity thresholds
• Remaining useful life (RUL) estimation from repeated measurements (linear, exponential, Kalman)
• Trend analysis and degradation onset detection
• Multi-signal trending and historical comparison
• Real-time streaming (MQTT/Kafka)
• Fleet dashboard for multi-asset monitoring
• CMMS integration (SAP, Maximo, Infor)

Ideas? Open a discussion or create an issue.

Are you using this?

I'd genuinely love to know. Whether you ran it on real machinery or just tried the sample data, drop a line in Discussions — one sentence about your machine or use case is enough. Real-world feedback directly shapes what gets built next.

Related

claude-stwinbox-diagnostics — Extends this project by connecting a physical edge sensor (STEVAL-STWINBX1) to Claude via MCP, with Claude Skills for guided condition monitoring. Same analysis engine, real hardware, operator-friendly reports.

Contributing

Contributions welcome from everyone — not just programmers. Domain experts, technical writers, and testers are equally valued. See CONTRIBUTING.md for paths tailored to your background.

Quick start: browse Issues for good first issue or help wanted labels.

Citation

@software{dimaggio_predictive_maintenance_mcp_2025,
  title   = {Predictive Maintenance MCP Server},
  author  = {Di Maggio, Luigi Gianpio},
  year    = {2025},
  version = {0.13.0},
  url     = {https://github.com/LGDiMaggio/predictive-maintenance-mcp},
  doi     = {10.5281/zenodo.17611542}
}

License

MIT — see LICENSE. Sample data is CC BY-NC-SA 4.0 (non-commercial); for commercial use, replace with your own machinery data.

Acknowledgments

MCP Python SDK (descended from FastMCP) · Model Context Protocol by Anthropic · Sample data from MathWorks · Core development assisted by Claude

An open-source predictive maintenance AI agent and condition monitoring copilot — built to support reliability engineers and the developer community.