NaeonAIr — AI Eye in the Sky for Crowd Safety

An AIoT-based drone system that analyzes crowd density, movement, and pressure in real time to prevent dangerous crowd situations.

NaeonAIr is a scalable smart-city safety platform that turns drones into intelligent airborne sensors, powered by edge AI, computer vision, and an IoT data backbone.

🌍 The Problem

Crowd disasters rarely happen without warning.

Before a crush incident occurs, there are measurable warning signs:

• Rising crowd density
• Conflicting movement flows
• Increasing physical pressure between groups

Unfortunately, ground-based cameras and human monitoring often fail to detect these patterns early enough.

We asked:

What if a network of AI-powered drones could monitor crowd conditions from above and warn authorities before danger escalates?

That idea became NaeonAIr — AI Eye in the Sky.

✨ What Is NaeonAIr?

NaeonAIr is an end-to-end AIoT ecosystem that combines:

🛸 Smart drone payloads

**🧠 **Edge AI processing (Jetson Nano gateway)

🤖 Crowd risk analysis server

🌐 Smart city IoT integration (oneM2M Mobius)

🖥 Real-time TypeScript monitoring dashboard

Together, they form a digital twin of crowd safety.

Device Prototype

🧩 System Overview

Drone Devices  →  Jetson Nano Gateway  →  AI Risk Server  →  Mobius IoT Platform  →  Web Dashboard

Each layer has a specific role in turning raw aerial footage into actionable safety intelligence.

🚁 1️⃣ Drone AI Payload (Edge Sensing Unit)

Each drone carries a lightweight AIoT payload module that transforms it into a flying sensor node.

🔧 Hardware Components

ESP32-CAM (OV2640) → Captures aerial images/video

Raspberry Pi Pico 2W → Controls GPS, servo, and communication

NEO-6M GPS → Provides geolocation tagging

F405 Flight Controller → Integrates with UAV power & telemetry

Servo Motor → Enables physical signaling experiments

LiPo Battery + ESC → Power system

The drone does not perform heavy AI processing — it focuses on data capture and transmission.

This circuit diagram demonstrates the ideal system of the payload integrated with the UAV circuitry

📡 What Each Drone Sends

Each drone transmits:

• Captured images
• GPS coordinates
• Device ID

These are sent to the ground gateway for centralized processing.

🧠 2️⃣ Jetson Nano Edge Gateway (Fleet Brain)

At the center of the system is an NVIDIA Jetson Nano, acting as a multi-device AI gateway.

This is what allows NaeonAIr to scale beyond a single drone.

🚦 Gateway Responsibilities

Device Hub → Manages communication with up to 100 drones simultaneously

Stream Router → Forwards live feeds to monitoring interface

AI Processing Node → Hosts crowd analysis services

IoT Bridge → Sends results to Mobius smart-city platform

Local Decision Layer → Can operate even with limited internet

Instead of every drone running AI, the Jetson Nano provides centralized edge intelligence.

🔄 Hybrid Data Flow Architecture

To balance real-time performance and scalable storage, NaeonAIr uses a hybrid communication model.

⚡ Real-Time Monitoring

• Gateway ↔ Dashboard via WebSocket
• Provides low-latency live monitoring

🌐 AI & IoT Integration

Gateway → AI Server : HTTP POST : Sends images for analysis

AI Server → Mobius : HTTP POST : Uploads risk results

Dashboard → Mobius : HTTP GET : Retrieves latest analytics

🧠 Efficient Media Handling

Instead of sending heavy image files to Mobius:

• AI server stores images locally
• Only image URLs + metadata are uploaded

This prevents overload of the IoT platform and keeps the system scalable.

🤖 3️⃣ AI Crowd Risk Analysis Server

This server turns drone images into crowd safety intelligence.

🔍 Step 1 — Person Detection

A fine-tuned YOLO model detects people from aerial imagery.

Output: Bounding boxes and crowd count.

🌊 Step 2 — Flow & Pressure Analysis

Unlike basic systems that only count people, NaeonAIr analyzes:

• Movement direction of crowd clusters
• Differences in flow between groups
• Zones where motion compresses into resistance

This produces a crowd pressure metric, a key factor in crush risks.

⚠️ Step 3 — Risk Score Calculation

The system combines:

• Crowd count
• Density trends
• Movement flow
• Pressure indicators

Into:

Risk Score (0–100)Risk Level: SAFE / CAUTION / DANGER

This provides an early warning system, not just statistics.

🌐 4️⃣ Smart City Integration — Mobius (oneM2M)

All analysis results are stored in Mobius oneM2M, enabling integration with city-scale systems.

📦 Mobius Data Structure

Mobius└── WisDrone    ├── crowd_analysis    ├── drone_status    └── alerts

This allows:

• Historical analysis
• Multi-system integration
• City-wide monitoring dashboards

🖥 5️⃣ Real-Time Monitoring Dashboard (TypeScript)

The human side of NaeonAIr is a TypeScript-based web dashboard.

This interface acts as a live control and monitoring center.

🎛 Dashboard Features

📦 Device Management

• Register and monitor multiple drones
• View active/inactive device status

🧠 AI Mode Selection

Operators can switch visualization modes:

ModeDescription

Density Shows crowd concentration

Danger Displays AI-estimated risk

📷 Live Camera Streams

Operators can view real-time drone video feeds.

🌍 3D Map-Based Visualization

Risk data is overlaid on a 3D city map, allowing:

• Spatial understanding of risk zones
• Tracking of crowd movement patterns
• Faster emergency response decisions

Not just numbers — but spatial risk awareness.

📈 Scalability & Future Expansion

Because the system is gateway-based, it can support:

• Up to 100 drone devices
• CCTV cameras
• Fixed smart sensors
• Mobile robots

All follow the same pipeline:

Device → Gateway → AI Analysis → Mobius → Dashboard

⚠️ Safety & Ethical Considerations

NaeonAIr is a research and safety-support system.

• Must comply with UAV flight regulations
• Must respect privacy laws
• Designed for crowd safety monitoring, not surveillance misuse
• Supports human decision-makers, not autonomous enforcement

🚀 Our Vision

NaeonAIr transforms drones from flying cameras into:

AI-powered, city-scale crowd safety sensors

By combining edge AI, computer vision, IoT integration, and real-time visualization, the system provides a digital twin of crowd risk — enabling faster, smarter, and safer event management.