[AfA2026-PhysicalAI] - SafeRide_AI

Arduino App Lab

connecting wires

Modulino™ Movement

Arduino® UNO™ Q 2GB

Mobile Phone

<html><body><h3>SafeRide AI – Gig Worker Safety Platform</h3>SafeRide AI is an AI-powered embedded safety system designed to protect two-wheeler gig delivery workers. Built on the Arduino UNO Q with Arduino App Lab, the platform combines motion sensing, lightweight AI, crash detection, SOS emergency support, and smartphone-based dashcam evidence logging. Millions of gig workers operate in unsafe road conditions and high-risk environments. SafeRide AI provides a low-cost, phone-first safety layer that can run directly on embedded hardware while leveraging the rider’s existing smartphone for location and video proof. 1. Project Description SafeRide AI is a real-time rider safety system that integrates embedded sensors with on-device AI to classify risky driving behavior, detect accidents, trigger SOS emergencies, and record incident evidence. The platform is specifically designed for gig delivery workers who often lack formal insurance protection, enabling safer mobility and stronger financial inclusion. 2. Hardware Lineup <figure><img src="https://projects.arduinocontent.cc/a53aa06a-79f5-4027-a3f0-9987f0729994.png"/><figcaption>System Architecture</figcaption></figure> SafeRide AI is built using affordable, scalable hardware components: <ol><li>Arduino UNO Q: Dual-core board responsible for real-time sensor acquisition and AI + dashboard hosting. </li><li>Modulino Movement Sensor (Accelerometer + Gyroscope): Captures inertial motion data used for driving pattern recognition and crash detection. </li><li>SOS Emergency Button: Physical panic trigger for robbery situations or urgent emergencies. </li><li>Smartphone Integration (Dashcam + GPS): Provides continuous video evidence, GPS tagging, and instant upload support for insurance-valid proof. </li></ol>Optional USB peripherals such as hubs, webcams, or SD storage can be added for extended standalone operation. 3. User Interface &amp; Feedback SafeRide AI provides rider-facing interaction through a built-in web dashboard accessible from any phone or laptop. The interface displays: <ol><li>Current driving pattern classification </li><li>Safety score (0–100) </li><li>Event timeline of harsh braking, sharp turns, turbulence, and crashes </li><li>SOS emergency alerts </li><li>Danger-zone location tagging </li></ol>The system also supports expandable physical feedback such as LEDs or buzzers for real-time rider warnings. 4. The AI Model SafeRide AI uses a lightweight Decision Tree classifier optimized for embedded deployment. The model processes accelerometer and gyroscope data locally at 62.5 Hz and classifies seven riding patterns, including normal riding, aggressive turns, harsh braking, road turbulence, and stationary theft-risk mode. Because inference runs entirely on-device with low latency (under 5 ms), the system provides instant safety responses without requiring cloud connectivity. 5. Software Architecture SafeRide AI leverages the Arduino UNO Q dual-core architecture: <ol><li>MCU Core (C++) handles real-time accelerometer/gyroscope sampling and SOS button monitoring. </li><li>Processor Core (Python – Arduino App Lab) performs feature extraction, Decision Tree inference, safety scoring, and dashboard hosting. </li><li>RPC Communication Bridge enables bidirectional messaging between hardware triggers and AI response logic. </li><li>Web Dashboard Layer provides live APIs and WebSocket-based updates for real-time rider monitoring and danger-zone visualization. </li></ol> 6. Visuals &amp; Media <h3>Video Link : <a href="https://drive.google.com/file/d/1-iFlsQWMyJ-xewqoKgR4exkIk_fIQAvc/view">https://drive.google.com/file/d/1-iFlsQWMyJ-xewqoKgR4exkIk_fIQAvc/view </a></h3>Screenshots : <figure><img src="https://projects.arduinocontent.cc/f26daf57-bc57-484f-9c04-c2cba03b8b36.png"/><figcaption>Main Dashboard</figcaption></figure> <figure><img src="https://projects.arduinocontent.cc/811f3891-c22a-4f29-9f8c-0d404a67311f.png"/><figcaption>User Analytics</figcaption></figure> <figure><img src="https://projects.arduinocontent.cc/bb4e0da2-082c-4673-aea0-d8d0098561e9.png"/><figcaption>Danger Map (Upcoming)</figcaption></figure> 7. Github Link : <a href="https://github.com/hrsh-kr/AfA2026-PhysicalAI-SafeRide_AI/tree/main">https://github.com/hrsh-kr/AfA2026-PhysicalAI-SafeRide_AI/tree/main </a> </body></html>

[AfA2026-PhysicalAI] - SafeRide_AI

SafeRide AI is an AI-powered embedded safety system for two-wheeler gig delivery workers, built on Arduino UNO Q with Arduino App Lab.