DATO
DATO: Decentralized Autonomous Traffic Orchestration The Problem: Modern smart-city traffic systems rely entirely on cloud connectivity.
Core trick
Cloud Gemini issues JSON macro-directives via function calls to an on-device Gemma node, which keeps running the last directive when connectivity drops.
Description
DATO: Decentralized Autonomous Traffic Orchestration
The Problem: Modern smart-city traffic systems rely entirely on cloud connectivity. If the internet goes down, intersections revert to stupid timers, causing gridlock and endangering emergency vehicles.
The Solution: DATO is an asymmetric Cloud-to-Edge AI system designed for 100% offline-resilient traffic management.
Instead of building a basic LLM chatbot, I engineered a true multi-agent architecture:
The Cloud Teacher (Strategic Agent): Powered by Gemini 3.5 Flash using the new Function Calling SDK. An operator can type natural language emergency commands (e.g., "Clear path for ambulance on North road"). Gemini autonomously calls tools to read the physical intersection sensors, formulates a strategy, and pushes a JSON macro-directive.
The Edge Student (Tactical Agent): A localized Gemma 4 model running directly at the intersection. It reads the Cloud's directive and implements a continuous Sense -> Decide -> Act -> Check loop.
Why it matters: Because heavy strategic reasoning happens in the cloud and tactical execution happens at the edge, the system is completely offline resilient. If the Wi-Fi is cut, the Gemma 4 Edge Node continues running the intersection safely and autonomously using local heuristic fallback loops and the last known directive.
README excerpt
# 🚦 Decentralized Autonomous Traffic Orchestration (DATO) An **Asymmetric Teacher-Student Multi-Agent** system for autonomous traffic management, built for the Google DeepMind x Cerebral Valley Bangalore Hackathon (July 2026). 🏆 **Track:** Cloud-to-Edge AI ## 🌟 The Magic: Emergency Preemption in Action When an ambulance gets stuck, my Cloud AI (Teacher) autonomously reads the intersection sensors, reasons about the traffic, and pushes an emergency strategy to the Edge AI (Student), which physically clears the cars in real-time. | Before: Massive 90-car jam on North corridor | After: Edge agent clears the North corridor | | :---: | :---: | | <img src="before_radar.png" width="400"/> | <img src="after_radar.png" width="400"/> | ## 🏗️ Architecture ```text ┌──────────────────────────────┐ directives.json ┌──────────────────────────────┐ │ CLOUD TEACHER │ ────────────────────────► │ EDGE STUDENT │ │ (Gemini 3.5 Flash) │ │ (Gemma 4 E2B — OFFLINE) │ │ │ │ │ │ • Natural language input │ │ • Reads directive locally │ │ • Autonomous Agent Loop │ │ • Runs Gemma 4 inference │ │ • Uses Function Calling │ │ • Stateful decision loop │ │ • Terminal Radar HUD │ │ • Error recov