AI Feature List

This table summarizes what PlantPulse AI can do, organized by module. AI is the explanation layer—it answers based on data collected by the Platform and clearly states when no supporting data is available.

How to Read This Page

• Where — 화면 indicates the copilot or admin console, API indicates REST, MCP indicates tool integration, and 설정 indicates environment variables.
• Services are divided into six categories—LLM, RAG, time series, vision, document understanding, and GPU—and the admin console provides the same information for each category: settings, API keys, endpoints, and logs.
• A PDF covering all four products is available under Downloads.

Questions and Answers (Copilot)

Feature	Description	Where	Details
AI Chat	Ask questions about the plant and receive answers with supporting evidence	UI	AI Chat Web
Evidence for This Answer	Review which tools were called, their arguments, and the returned results	UI	AI Chat Web
Conversation Search	Find previous conversations by title	UI	AI Chat Web
Screen Map	Shows which screen answers which types of questions	Reference	Copilot Screen Map
Anomaly Detection and Incident Response	Review detected anomalies, acknowledge them, and suppress them	UI	Incident Response
Daily Action Recommendations, Reports, and Weekly Briefings	Receive today's recommended actions and review the results and the past week	UI	Copilot Screen Map
Mobile	Ask questions from a phone in the field	UI	Copilot Screen Map
Public API	Allows external systems to call chat, daily action recommendations, and the tool list (X-API-Key)	API	API Integration Guide
Workflow	Who does what and when	Reference	Business Process

Models (AURA)

Feature	Description	Where	Details
Diagnostic Model	Answers industrial questions using an internally trained model	API · UI	AURA Models
Retrieval Model	Finds supporting evidence in documents using embeddings	API	AURA Models
Precision Ranking Model	Selects the best answer from the candidates using a reranker	API	AURA Models
On-Premises Serving	Keeps models and data within the organization	Settings	AURA Technical White Paper
Model Management	View deployed models and how to call them	UI · Settings	AI Admin Console
Playground	Test models directly from the UI	UI	AI Admin Console

Document Retrieval (RAG)

Feature	Description	Where	Details
Document Indexing	Reads and indexes manuals and procedures	UI · API	RAG
Evidence Retrieval	Finds passages relevant to a question	UI · API · MCP	RAG
Synchronization History	View what was added and when	UI	AI Admin Console
Ontology	Connects relationships in a plant knowledge graph	API	Ontology

Time Series Analysis (TimeSeries-Insight)

Feature	Description	Where	Details
Anomaly Detection	Equipment, systems, OEE, RAM, EMS, work orders, and alarms	UI · API · MCP	TimeSeries-Insight
Forecasting	Forecasts future trends for the same seven target types	UI · API · MCP	TimeSeries-Insight
Equipment Health Index	Represents equipment condition as a single score	UI · MCP	TimeSeries-Insight
In-Depth Alarm Analysis	Correlation, duration, frequency, and forecasting	UI · MCP	TimeSeries-Insight
Feature Store	Precomputes values for faster responses	UI · API	TimeSeries-Insight
33 Analysis Tools	List of tools exposed through MCP	MCP	Time Series Insight Tools
Model Class Configuration	Selects the loader that matches the checkpoint generation	Settings	TimeSeries-Insight

Vision and Document Understanding

Feature	Description	Where	Details
Vision	Reads and describes images	API · UI	AI Admin Console
Document Understanding (Parser)	Extracts content from PDFs and drawings	API · UI	AI Admin Console

Real-Time Integration (MCP)

Feature	Description	Where	Details
Four MCP Servers	Allow assistants to use AI services as tools	MCP	MCP Servers
Tool Map	Shows what each tool does	Reference	MCP Tool Map
Platform Integration	Reads equipment, tags, and alarms in real time	MCP	MCP Servers
Tool Approval Gate and Invocation Ledger	Prevents state-changing tools from running in chat and records every invocation	Settings	MCP Servers

Administration and Operations

Feature	Description	Where	Details
Admin Console	Manages all six service categories from one UI	UI	AI Admin Console
Copilot Administration	Manages users, conversations, and policies	UI	Copilot Admin Console
Public API Keys	Issues and revokes keys for copilot /api/v1 and stores them only as hashes	UI	Copilot Admin Console
API Keys	Issues and revokes keys for each service category	UI	AI Admin Console
Endpoint List	Shows what to call and where	UI · Reference	API Integration Guide
Logs	View logs for each service category in the UI	UI	Logs and Monitoring
GPU Status	Shows how many GPUs are available and what is deployed on them	UI	Logs and Monitoring
Database Management	Shows what data is stored and where	Settings	Database Management
Service Operations	Start, stop, and recover services	CLI · Settings	Service Operations
Credential Rotation	Change passwords and API keys	Settings	Change Passwords and API Keys
Port Map	Shows which port each service uses	Reference	Service Port Map

Installation

Feature	Description	Where	Details
One-Line Installation	Installs with a single command	CLI	One-Line Installation
Air-Gapped Installation	Installs without internet access	CLI	Air-Gapped Installation
System Requirements	GPU, memory, and disk requirements	Reference	Requirements

All Four Products in One Guide

Language	Pages	File
Korean	9 pages · landscape	PlantPulse Feature List (PDF)
English	12 pages · landscape	PlantPulse Feature List (PDF)

This page is the authoritative source—the PDF is generated by a script that compiles the feature lists for all four products, so its content is always identical.

See Also

Overall Structure · Downloads · AURA Technical White Paper · Architecture · User Guide