Search

Cookies

We use cookies to improve your experience. By continuing, you accept our use of cookies.

Technology

Bengaluru Engineer's AI System Detects Potholes, Files Complaints Automatically

· · 2 min read

Bengaluru-based engineer Gaurav Sen developed an AI-powered system that uses a dashcam to detect potholes, identify responsible contractors, and automatically generate official complaints with location data and photos. This aims to streamline road defect reporting.

Bengaluru, India – A local engineer has developed an innovative artificial intelligence system designed to tackle the city's persistent pothole problem. Gaurav Sen, founder of InterviewReady, created the AI-powered solution after experiencing vehicle damage due to poor road conditions.

How the AI Pothole Detection System Works

Sen's system utilizes a dashcam equipped with GPS and an accelerometer, which records road conditions as he drives. An integrated AI vision model then analyzes this data to detect potholes, classifying them by size into small, medium, or large categories. The system is also designed to differentiate between actual potholes and other road features like speed breakers, ensuring accurate identification without requiring the driver to stop.

Automating Complaints and Identifying Responsibility

Beyond simple detection, the AI system takes a crucial step further. It cross-references approximately 2,900 government contracts to identify the specific tender linked to the damaged road. This allows the system to pinpoint the contractor responsible for the road's maintenance and the relevant government official overseeing the work.

Many road contracts include provisions for maintenance and defect liability, meaning contractors are often responsible for addressing issues within a specified period. By linking defects directly to responsible parties, the AI system aims to shift accountability and expedite repairs.

Streamlining the Complaint Process

Once a pothole is detected and responsibility is identified, the system automatically generates a structured complaint. This complaint includes photographic evidence of the pothole, its precise geographical coordinates, and details of the responsible contractor and official, making it ready for immediate submission to civic authorities.

In a demonstration, the system successfully detected 12 potholes during a single drive to work and generated 12 corresponding complaints. While Bengaluru already has existing mechanisms for reporting potholes, such as the BBMP’s “Fix My Street” and the Greater Bengaluru Authority’s “Fix Pothole” services, Sen's AI system adds a layer of efficiency by directly connecting road defects with those accountable for their repair.

Related