Analyze industrial vibration data and diagnose bearing faults using ISO 20816 standards and machine learning anomaly detection. Built for maintenance engineers and DevOps teams integrating predictive maintenance into monitoring systems.
io.github.LGDiMaggio/predictive-maintenance-mcp
Local install
STDIO
No auth required
How models use it and what it is built for.
Analyze industrial vibration data and diagnose bearing faults using ISO 20816 standards and machine learning anomaly detection. Built for maintenance engineers and DevOps teams integrating predictive maintenance into monitoring systems.
Local install — runs as a subprocess.
Where to find authoritative docs and source for Predictive Maintenance.
Paste any of these into Agent Studio after connecting Predictive Maintenance.
Common questions about connecting and running Predictive Maintenance.
What vibration standards does this server support?
The server implements ISO 20816 for vibration severity assessment and bearing fault diagnosis. It uses machine learning anomaly detection to identify deviations from normal operating patterns.
How do I install and run predictive-maintenance-mcp?
Install via `uvx predictive-maintenance-mcp@0.8.0`. The server runs over stdio transport, making it compatible with standard MCP client integrations and Claude Desktop.
What data format does the server expect for vibration analysis?
The server accepts industrial vibration sensor data. Refer to the official repository documentation for specific schema requirements and supported input formats for bearing diagnostics.
Can this replace my existing condition monitoring system?
This MCP server is designed as an integration layer for predictive maintenance workflows, not a standalone SCADA replacement. It works best alongside existing monitoring infrastructure to add ML-driven anomaly detection capabilities.
What are the computational requirements for anomaly detection?
The server runs as a lightweight stdio process. For large-scale industrial deployments, test with your typical vibration data volume to ensure performance meets your monitoring interval requirements.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent Studio