6. OSA-Web: Fleet Monitoring Dashboard

6.1. Overview

The OSA-Web dashboard is a comprehensive Django-based application designed for multi-camera fleet monitoring. It utilizes real-time WebSocket frame streaming and a Celery task queue to perform AI-powered shelf analysis asynchronously, allowing centralized management of entire retail spaces.

6.2. Features

Multi-Camera Management: Easily configure and monitor numerous RTSP streams simultaneously.
Real-time WebSockets: Streams annotated frames and live analytics directly to the browser with ultra-low latency.
Distributed Processing: Utilizes Redis and Celery to execute heavy computer-vision pipelines asynchronously.
AI Assistant Co-Pilot: Integrated intelligent assistant capable of understanding natural language queries and generating dynamic charts based on inventory data.

6.3. User Interface

6.3.1. Home Dashboard

The Web Dashboard Home provides a centralized view of all active monitoring sessions, fleet-wide alerts, and top-level KPIs.

6.3.2. Real-Time Monitoring

Watch a live, annotated feed of any active camera session with associated real-time metrics streaming alongside the video.

6.3.3. Configuration

Administrators can easily configure camera endpoints, model paths, and alert rules through the management interface.

6.4. Session Management

6.4.1. Sessions History

Browse past monitoring sessions, review performance data, and access historical analytics.

6.4.2. Session Details

Examine a specific session in depth, exploring the timeline of events, threshold triggers, and overall performance during that period.

6.5. Analytics & Reporting

6.5.1. KPIs Dashboard

Track aggregated Key Performance Indicators across the entire store or drill down into specific camera zones.

6.5.2. Stock Inventory

View current stock levels, missing items, and capacity data for all monitored product categories.

6.5.3. Temporal Analysis

Analyze stock fluctuations and restock cycles over time using interactive temporal charts.

6.6. AI Assistant Co-Pilot

The OSA-Web dashboard includes a powerful Groq-powered AI Assistant.

6.6.1. Interactive Queries

Ask natural language questions about your inventory data (e.g., « Which products are critically low? », « Show me the restocking trends for Cola »).

6.6.2. Dynamic Graph Generation

The AI Assistant can generate dynamic, interactive graphs directly in the chat interface in response to complex analytical questions.