Tired of potholes? Bengaluru techie builds AI-powered app that detects damaged roads and identifies contractors

A Bengaluru engineer has turned his daily frustration with potholes into an AI-powered system that can detect damaged roads, locate them precisely and identify the contractor responsible for the road. The system combines a dashcam, GPS, an accelerometer and AI vision models, while also checking 2,900 government contracts to connect each pothole with the relevant road project.

From bumpy roads to an AI solution

For many motorists, hitting a pothole means a rough ride, possible vehicle damage and another complaint to the civic authorities. For Bengaluru-based engineer Gaurav Sen, it became a problem that could be tackled with technology.

Also Read: Man fails to get any suitable marriage match from matrimonial website: Paid Rs 25,900 for 100 contacts but received only 51; consumer commission orders Rs 20,000 payout


Sen demonstrated his system in a video shared on Instagram in collaboration with ChatGPT India. The video later reached X, where it drew attention for its unusual approach to a familiar problem on Indian roads.

Instead of simply recording potholes, Sen's system attempts to build a complete record of each road defect.

How the pothole detection system works

Sen installed a dashcam, GPS and an accelerometer in his car. As he drives, the equipment collects road-related information along his route.

The recorded footage is then examined using AI vision models. The system can identify potholes and classify them according to their size.


The technology is also designed to deal with a common problem on Indian roads: potholes that can resemble speed breakers when viewed from a distance.

Sen said he used Codex to develop an app that collects the relevant information and sends it to OpenAI for processing.

AI goes beyond finding potholes

Finding a pothole is only part of the problem. The bigger question is often who should repair it. Sen's system tackles that by searching through 2,900 government contracts. It attempts to identify the contractor responsible for the particular stretch of road where the pothole has been detected.

This can be especially useful when a road remains under warranty. In such cases, the contractor may be responsible for fixing defects without the government having to bear an additional repair cost.

Four seconds to create a complaint record

The app can compile several pieces of information into a structured record within about four seconds. The record includes the pothole photograph, its exact location, the relevant tender number and the officer responsible for the road.

That turns a simple drive into a collection of documented evidence that can potentially be used while raising a complaint with civic authorities.

One commute found 12 potholes

Sen also tested the system during one of his regular drives to work. The app detected 12 potholes on the route, providing photographic and location-based evidence for each one. The system could then be used to prepare complaints linked to the respective road authorities or contractors.

Also Read: Railway passengers take note: If you drink water from station coolers, CAG has found E coli in samples from 8 locations

The experiment showed how AI can potentially reduce the effort involved in documenting everyday civic problems.

Social media users take notice

The project gained wider attention after the video reached X. Users praised the idea of applying AI to a problem that affects motorists in Bengaluru and many other Indian cities.

The social media post explained the system this way: "An Indian engineer in Bengaluru got tired of potholes destroying his car. So he built an app with Codex that detects them, finds who is responsible, and files a complaint. Here is how it works. He installed a dashcam with GPS and an accelerometer. The app records everything while he drives. A vision model classifies every pothole by size. Small, medium, large. But here is the clever part. Most Indian roads are under warranty. The contractor who built them is legally responsible for fixing potholes for free. His app searches through 2,900 government contracts, finds the exact tender number, identifies the officer responsible, attaches the photo and geolocation, and generates a ready-to-file complaint. One drive to work. 12 potholes detected. 12 complaints ready."

Some users also suggested that civic agencies could adopt similar technology to identify damaged roads and act on complaints more quickly.

A different way to report civic problems

Sen's experiment shows how a routine commute can generate useful civic data when everyday hardware is combined with AI.

Instead of stopping at identifying potholes, the system attempts to connect each road defect with the contract, contractor and official associated with that stretch. What started as an attempt to make one Bengaluru commute less bumpy has turned into a broader experiment in making road complaints more detailed and easier to act on.