AI tools, scored for your job
Get paid to train AI
🤝 You're in the Free & Open Source library — community-built tools you can use without paying. A different concept from our scored AI directory.
Home › Free & Open Source › Hospitality › predictive-maintenance-mcp
✦ Free & open sourceHospitality

predictive-maintenance-mcp

An MCP server that gives AI assistants evidence-based vibration diagnostics and ISO-cited machinery fault reports.

AI vibration diagnostics for heavy facility machinery

Turn your AI desktop assistant into a condition monitoring aid for pumps, motors, and bearing diagnostics.

Free forever · open source · MIT
✓ Needs IT✓ Browser-based✓ Real-time collaboration
6.6
ZEKAI
Evidence-based AI maintenance
Free & open source · Hospitality
Ease of setup4
Usefulness8
Maintenance9.5
Community5
★ 106
GITHUB
Oct 2026
UPDATED
MIT
LICENSE
🛠 Don't want to self-host? We can set it up →
⚡ Quick answer

This tool is a Python-based Model Context Protocol (MCP) server that connects LLM clients like Claude Desktop to local vibration analysis engines. It suits facility maintenance leads and engineers in hospitality infrastructure who want AI-assisted equipment fault detection anchored strictly in ISO standards without sending sensor data to third-party clouds.

Runs on
Cross-platform
Hosting
Desktop
Languages
Python
License
MIT

What it is

It exposes 42 specialized MCP endpoints that execute envelope analysis, calculate bearing fault frequencies, and assess ISO 20816-3 severity zones. Instead of letting LLMs guess root causes from filenames or statistical summaries alone, the server forces the AI to present matching spectral evidence and outputs interactive HTML diagnostic reports.

Who it's for

It suits technical facility managers, reliability engineers, and operations teams managing pumps, HVAC systems, or guest-service machinery who already use desktop AI assistants. It is not suitable for non-technical staff expecting a turn-key web UI, as it requires configuring local Python environments and preparing single-column CSV signal files.

Key features

Strict Evidence Matching

Requires spectral line evidence above noise floors before confirming fault indications rather than letting the LLM guess.

ISO 20816-3 Severity Assessment

Evaluates vibration signals directly against ISO standard velocity thresholds to assign machine severity zones.

Interactive HTML Reports

Generates self-contained diagnostic report files featuring annotated spectral charts and server-authored evaluative text.

Local Infrastructure Security

Runs completely on your local machine so raw vibration data and PDF equipment manuals never leave your network.

Anomaly Baseline Training

Trains statistical anomaly detectors on healthy operational baselines to flag potential outliers in newly acquired signals.

How to start

Up and running in 3 steps

1
Clone the repository and set up the Python 3.11 environment using the uv package manager.
2
Run the provided configuration script or add the MCP server executable path into your Claude Desktop JSON config.
3
Place your single-column CSV signals with companion JSON metadata files into the signals directory and prompt your AI client to run envelope analysis.
Real talk

Our honest verdict

What's great

  • Refuses to let the LLM hallucinate or infer diagnoses from filenames alone.
  • Cites ISO 20816-3 standards directly for machinery severity assessments.
  • Keeps raw sensor data completely on your local infrastructure without external network calls.
  • Provides 42 dedicated MCP endpoints and 8 diagnostic skills for Claude integrations.

Where it falls short

  • Requires installing Python 3.11, the `uv` tool, and editing JSON configuration files manually.
  • Requires signal files formatted as single-column CSVs paired with custom JSON metadata files.
  • Does not include a standalone graphical interface, relying on third-party MCP clients.
  • Demands working knowledge of vibration frequencies, shaft RPM, and envelope spectra.
Cost

Pricing & license

Free under the MIT license. The only cost is running it on your own local hardware and paying for any underlying API usage required by your chosen MCP client.

Paid work

Open remote AI-training roles

873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.

See all open roles →
Compare

Alternatives

Get more

Use it, or have us set it up

📄 Free PDF: predictive-maintenance-mcp for Hospitality
A plain-language guide to getting your first result. Sent to your inbox.
🔔 64 people want a full video tutorial — add your vote
🛠 Don't want to set it up yourself?
We'll install, host & configure predictive-maintenance-mcp for you — and train your team.
Get it done for you ↗
Questions

FAQ

Does this tool send my facility's vibration data to external cloud servers?

No, the MCP server runs entirely on your local machine and processes all signal files locally.

Can it read equipment manuals directly?

Yes, you can instruct your AI client to read PDF manuals, extract bearing model specs, calculate expected fault lines, and test them in the server.

Radar

Latest AI news

✉️ The weekly free toolkit
The best new free & open-source tools for Hospitality — one short email a week, in your language.
Free · 1 email/week · unsubscribe anytime
Get paid to train AI →

See Zekai first in Google