PlantPulse AI Platform
If you're getting started, follow this order
- Make the chatbot answer from our docs — includes a public demo and video
- Example questions — what you can ask
- One-line installation → Change initial password
- Work process — how to keep it useful after deployment
"All factory data, for everyone, one natural language question away"
info
Online manual: https://kopens.gitbook.io/plantpulse-ai
PlantPulse AI Platform is an industrial AI copilot platform that integrates real-time field data, AI analytics, and knowledge management for manufacturing operations.
When factory operators ask questions in natural language, the AI automatically performs real-time sensor data queries, anomaly detection, predictive analytics, and technical document search — enabling immediate and accurate decision-making.
Key figures
| Item | Value |
|---|---|
| Integrated MCP data tools | 116 (read-only) |
| TimeSeries-Insight analytics tools | 27 |
| Knowledge graph (ontology) tools | 5 |
| RAG search tools | 3 |
| Ontology node types | 14 |
Platform architecture
| Component | Role |
|---|---|
| AI Chat Web | User interface — natural language conversational AI copilot (built-in agent) |
| Integrated MCP | Real-time factory data gateway (Platform server-web /api/v5/mcp, includes ontology) |
| Ontology | Factory knowledge graph — equipment relationship/structure analysis (included in Integrated MCP) |
| TimeSeries-Insight | AI anomaly detection + forecasting + feature store (27 analytics tools) |
| RAG | LightRAG-based technical document search — manuals, SOPs, regulations at your fingertips |
| AURA | Manufacturing-specialized LLM — domain-fine-tuned model for industry |
Adoption impact
| Area | Before (Legacy) | After (PlantPulse AI) |
|---|---|---|
| Status overview | SCADA + spreadsheets + manuals 30 min–1 hour | Natural language question 10 seconds |
| Anomaly detection | Rule-based alarms (reactive) | AI early warning + root cause estimate |
| Maintenance approach | Preventive (calendar-based) | Predictive maintenance (condition-based) |
| Document search | Manual file server exploration | AI hybrid search instant |
| AI cost | Cloud API pay-per-use | On-premises fixed cost unlimited |
| Data security | External cloud transmission | On-premises kept complete |
Technology stack
| Layer | Technology |
|---|---|
| AI models | AURA LLM, IBM Granite TSPulse, TinyTimeMixer TTM |
| AI framework | MCP (Model Context Protocol), LiteLLM, LightRAG |
| Backend | Java 21 / Spring Boot 3.4, Python / FastAPI |
| Graph DB | Neo4j |
| Vector DB | Qdrant |
| Relational DB | PostgreSQL 18 |
| Time series DB | Apache Cassandra |
| Cache | Valkey (Redis compatible) |
| Object storage | MinIO (S3 compatible) |
| GPU | NVIDIA DGX Spark compatible (low-spec operation) |
| Deployment | On-premises / Edge / Cloud |
Target industries
| Industry | Key use cases |
|---|---|
| Food & beverage manufacturing | Production line OEE management, quality monitoring, energy optimization |
| Chemical & petrochemical | Early process anomaly detection, safety compliance, equipment health management |
| Automotive parts | Predictive maintenance, work instruction quality tracking, line balancing |
| Semiconductor & electronics | Fine-grain anomaly detection, yield forecasting, cleanroom environment monitoring |
| Steel & metals | High-temperature equipment monitoring, energy management, maintenance cycle optimization |