An App That Detects Potholes? Bengaluru Engineer Gaurav Sen Builds AI Tool That Also Identifies Contractors Responsible For Fixing Them
Bengaluru engineer Gaurav Sen has allegedly built an AI-powered system that detects potholes from dashcam footage and maps them to government road contracts. Using GPS and an accelerometer, the system identifies potholes and matches their locations with around 2,900 contracts to reveal the contractor, tender number and government officer overseeing the road work.

X/ Indian Tech & Infra
A Bengaluru-based engineer, Gaurav Sen, has allegedly built an AI-powered system that detects potholes using dashcam footage and cross-references their locations with government road contracts to identify the contractor and officer responsible for repairs.
How the system works?
Sen equipped his car with a dashcam, a GPS unit and an accelerometer to capture road footage during regular drives. The footage is analysed using AI vision models that identify potholes and classify them by size, according to his post on Instagram. The system then cross-references the pothole's location against a database of around 2,900 government road contracts to identify the contractor responsible for that stretch of road, along with the relevant tender number and the government officer overseeing it.
Linking defects to contractual accountability
According to the post, the system's underlying premise is that many Indian roads remain under contractual warranty, during which the contractor who built the road is required to fix defects such as potholes without additional cost to the government. By connecting a detected pothole to the applicable tender and contractor, the tool is designed to generate a complaint that references the specific accountability framework already in place, rather than relying on general civic grievance reporting.
Complaint generation and real-world test
Each detected pothole is compiled into a record that includes a geo-tagged photograph, the location, the tender number and the responsible officer, generating what the post described as a ready-to-file complaint. In a test run during a single commute, the system detected 12 potholes and generated 12 corresponding complaints, according to the post.
Wider civic-tech implications
The demonstration drew attention for extending beyond individual complaint filing toward a broader accountability mechanism for road maintenance. Civic authorities could potentially adopt similar AI-based systems to survey road conditions and identify damaged stretches at scale, rather than relying solely on complaints filed by individual commuters.
RECENT STORIES
-
'Gujarat-Centric Development Needs To Change': Ex-Tamil Nadu BJP Chief Annamalai Criticises Centre... -
An App That Detects Potholes? Bengaluru Engineer Gaurav Sen Builds AI Tool That Also Identifies... -
Govt To Announce High-Level Committee On Banks This Month, Says Nirmala Sitharaman -
Supreme Court Grants Interim Relief To Rahul Gandhi In Disproportionate Assets Case, Stays Allahabad... -
WCL Season 3 Set To Begin October 3 With David Warner, Yuvraj Singh, AB de Villiers & Star-Studded...
