Capabilities

Everything You Need for Machine Diagnostics

Professional-grade tools, accessible through any MCP-compatible AI assistant.

📊

FFT Spectrum Analysis

Frequency-domain decomposition with automatic peak detection, harmonic identification, and dB-normalized visualization.

🔍

Envelope Analysis

Bearing-fault-focused demodulation. Reveals characteristic defect frequencies (BPFO, BPFI, BSF, FTF) hidden in raw signals.

📏

ISO 20816-3 Assessment

Automated vibration severity evaluation per international standards, with unit conversion and hypothesis-based confirmation flow.

🤖

ML Anomaly Detection

Train on healthy baselines, detect anomalies in new signals. OneClassSVM & LocalOutlierFactor with automatic feature extraction.

📄

Interactive HTML Reports

Publication-quality reports with Plotly charts, auto-generated summaries. Shareable files for ops teams and management.

📁

Multi-Format Ingestion

Load from CSV, MATLAB (.mat), WAV, NumPy (.npy), and Parquet. Unified loading with automatic metadata handling.

Workflow

Three Steps to Expert Diagnostics

From install to actionable insight in under five minutes.

Install & Connect

One command: pip install predictive-maintenance-mcp. Add to Claude Desktop, VS Code, or any MCP client.

Ask in Plain Language

"Analyze the bearing vibration data and check if there's an outer race fault."

Get Expert Analysis

AI orchestrates FFT, envelope, ISO checks — delivers evidence-based results with interactive reports.

Live Preview

What a Conversation Looks Like

Real examples of how maintenance teams interact with the assistant.

Claude Desktop — PM-MCP Connected

You

Is this pump safe to run for the next shift?

Assistant

I ran an ISO 20816-3 assessment on the latest vibration data. The RMS velocity is 4.2 mm/s, placing this pump in Zone C (Alert). I recommend planning inspection within 24 hours and reducing load by ~15%.

You

What's the most likely fault? Create a report for my supervisor.

Assistant

Envelope analysis shows peaks at BPFO harmonics (1x, 2x, 3x), consistent with possible outer race wear on the drive-end bearing. I've generated an HTML report with spectrum, envelope, and ISO charts — saved to reports/.

Use Cases

Built for Two Audiences

Whether you maintain machines or build AI tools — this is your starting point.

🔧

Reliability & Maintenance Engineers

Diagnose machine health through natural language. No signal processing PhD required.

Screen machine health conversationally
Identify likely fault families from real data
Generate shareable reports for management
Follow guided diagnostic workflows

Engineer Quick Start →

💻

AI & Software Developers

Learn MCP tool design and build industrial copilots on a production-ready foundation.

Study real MCP tool architecture patterns
Extend with custom tools and datasets
Build reusable industrial AI workflows
Use Copilot Skills for quick integration

Developer Quick Start →

Gallery

Reports & Visualizations

Professional outputs generated through natural language commands.

Envelope analysis showing BPFO, BPFI, BSF bearing fault frequencies — Envelope analysis — bearing defect frequency identification

ISO 20816-3 vibration severity chart with zone classification — ISO 20816-3 vibration severity assessment

MCP server architecture showing 25 tools, 4 resources, 4 prompts — MCP server architecture overview

Foundation

Technology Stack

Built on battle-tested Python libraries and the open Model Context Protocol.

⚡

FastMCP

High-performance MCP server with auto schema generation and transport handling.

🧪

SciPy + NumPy

Industry-standard scientific computing for signal processing and spectral analysis.

🧠

scikit-learn

ML anomaly detection with OneClassSVM & LocalOutlierFactor, feature extraction on healthy baselines.