🚗 AutoMate: AI-Powered Vehicle Diagnostics & Cost Estimator AutoMate is an intelligent vehicle diagnostic dashboard that combines Computer Vision with a Deterministic Expert System to help drivers identify warning lights, understand safety risks, and estimate repair costs in real-time.

Designed specifically for the Sri Lankan context, it provides localized cost estimates (LKR) and tracks recurring vehicle issues to prevent breakdowns.

🌟 Key Features 👁️ AI-Powered Recognition: Uses a fine-tuned MobileNetV2 Neural Network to instantly identify dashboard symbols (Check Engine, Battery, Oil Pressure, etc.) from user-uploaded photos.

🛡️ Safety-First Advice: Implements a Deterministic Retrieval System to provide manufacturer-approved safety instructions, ensuring 100% consistent and safe advice.

🇱🇰 Localized Cost Estimation: Automatically calculates repair costs and labor hours based on current Sri Lankan market rates for parts and workshop labor.

🧠 Intelligent History Tracking: An integrated History Agent (PostgreSQL) detects if a specific error is recurring, escalating the urgency level for the user.

💬 Natural Language Support: Users can simply type "my brakes feel weird," and the NLP engine will map it to the correct technical issue.

🏗️ Technical Architecture This project utilizes a Hybrid Neuro-Symbolic Architecture:

Perception Layer (Probabilistic): A Deep Learning model (MobileNetV2) handles the uncertainty of visual recognition (classifying images).

Decision Layer (Deterministic): A Rule-Based engine handles the logic, safety advice, and cost calculations to ensure accuracy and reliability.

🛠️ Tech Stack Frontend: Python (Streamlit) with custom Glassmorphism CSS.

AI & Vision: TensorFlow, Keras, MobileNetV2 (Transfer Learning).

Data Processing: Pandas (CSV Knowledge Base), NumPy.

Database: PostgreSQL (Neon Tech) for user auth and history logs.

Deployment: Streamlit Cloud / Localhost.

🚀 Usage Log In: Secure authentication to save your vehicle history.

Diagnose:

Option A: Upload a photo of your dashboard.

Option B: Type the symptom (e.g., "car is overheating").

Report: View the identified issue, immediate safety actions, estimated repair cost (Rs.), and labor time.

🔮 Future Scope Object Detection (YOLO): To detect multiple warning lights in a single image simultaneously.

Mechanic Integration: Connect users directly with nearby workshops via Google Maps API.

Voice Assistant: Voice-activated diagnostics for hands-free usage while driving.