OffPI

• 

• 

• 

• 

• 

Inspiration

During wars, natural disasters, and infrastructure failures, communication is often the first system to collapse. Without internet, electricity, or cellular coverage, people lose the ability to request help, and responders lose critical situational awareness.

We were inspired by real crisis scenarios where victims were physically safe enough to survive — but digitally isolated.

What it does

OffPI is an AI-powered offline emergency communication system designed for blackout and network failure situations.

It enables:

• SOS alerts without internet • Device-to-device mesh propagation • Voice-triggered emergency broadcasting • AI-based prioritization & filtering • Raspberry Pi-powered local crisis hubs • LoRa long-range fallback connectivity

Even when networks disappear, OffPI keeps critical information moving.

How we built it

🧰 How we built it

We combined software + edge computing + AI + resilient networking:

📱 Mobile App

Swift / SwiftUI

Multipeer Connectivity (peer-to-peer mesh)

Low-power broadcasting logic

Voice shortcuts

Emergency medical profile

🧰 Edge Node

Raspberry Pi

Local dashboard hosting

Store-and-forward buffering

Offline AI inference

📡 Connectivity Layer

LoRa modules for long-range offline transmission

🤖 AI Layer

Emergency classification

Priority scoring

Duplicate/noise filtering

Challenges we ran into

• Designing reliable offline message propagation • Preventing broadcast storms & loops • Power/battery optimization • Simulating crisis-like connectivity conditions • Balancing AI capability with edge hardware limits

Accomplishments that we're proud of

✅ Built a functional offline communication concept ✅ Designed mesh-based alert propagation ✅ Integrated AI decision support logic ✅ Created Raspberry Pi crisis hub architecture ✅ Developed realistic emergency use-case scenarios ✅ Delivered a socially impactful solution

What we learned

• Offline-first design is fundamentally different from cloud apps • Edge AI must be lightweight & efficient • Mesh networks require careful relay control • Crisis UX must be extremely simple • Resilience > Features during emergencies

What's next for OffPI

• Field testing with real devices • Enhanced mesh location estimation • Stronger AI models at the edge • Android version • Hardware miniaturization • Deployment kits for emergency vehicles & institutions

Built With

• lora
• node.js
• raspberry-pi
• swif