Automotiq

AI-Powered Vehicle Diagnostics System (2025)

Project Overview

Automotiq is a comprehensive vehicle diagnostic system that utilizes on-device AI inference to provide drivers with intelligent vehicle health monitoring and troubleshooting. The app leverages multimodal AI models to analyze diagnostic trouble codes (DTCs) streamed from an OBD2 dongle via Bluetooth Low Energy (BLE) and provides insights on vehicle faults as well as repair guidance.

Key Features

  • AI-Powered Diagnostics: Advanced analysis of diagnostic trouble codes using Gemma 3n
  • Multimodal Chat Interface: Text and image-based AI assistance for automotive issues
  • OBD2 Integration: Real-time vehicle data collection via Bluetooth Low Energy
  • Agentic Solution: Tests different candidate CAN commands to solve software-related problems
  • Offline Capability: Local AI processing and diagnostic database

Technical Implementation

The system combines cutting-edge AI technology with automotive diagnostics:

  • On-Device AI: Local processing ensures privacy and offline functionality
  • BLE Communication: Seamless data streaming from OBD2 devices
  • Multimodal AI: Handles both text queries and image analysis for comprehensive diagnostics
  • CAN Bus Integration: Direct communication with vehicle systems for advanced troubleshooting

Pitch Deck

Recognition

Grand Prize Winner - Google DeepMind’s AI Agent Hackathon

Automotiq won the Grand Prize in Google DeepMind's AI Agent Hackathon

Impact

Automotiq addresses a critical need in automotive maintenance by:

  • Democratizing vehicle diagnostics for everyday drivers
  • Reducing dependency on expensive diagnostic tools
  • Providing intelligent, context-aware repair guidance
  • Enabling proactive vehicle maintenance through AI insights

Development Timeline

  • May 2025: Project development and implementation
  • May 2025: Google DeepMind AI Agent Hackathon participation and victory

This project represents a significant advancement in the intersection of AI technology and automotive diagnostics, showcasing the potential for intelligent systems to transform traditional industries.