A sudden jolt on a Bengaluru road is usually enough to tell a motorist that another pothole has entered the journey. But instead of simply driving past it or adding another complaint to the long list of civic grievances, Bengaluru-based engineer Gaurav Sen decided to ask a more useful question: Can technology identify the pothole and also identify the person responsible for fixing it?That question has produced an intriguing experiment involving a dashcam, GPS, an accelerometer, artificial intelligence, and a database containing thousands of government road contracts.Sen’s system does not merely spot damaged stretches of road. It attempts to connect each pothole with the relevant contractor, tender, and government official, potentially turning a routine road complaint into a documented trail of accountability.
instagram.com/reel/DbqG442NwiQ/?utm_source=ig_embed&ig_rid=AnthGdAKYaIXVDd9MkNsbmL
The idea began with a familiar Bengaluru problem
Bengaluru’s roads have long been a source of frustration for commuters. A pothole can appear as little more than a depression in the asphalt, yet hitting one at speed can mean anything from an uncomfortable ride to damage to tires, wheels, or suspension.Sen approached the problem differently. A dashcam mounted in his car records the road while GPS captures the vehicle’s location and an accelerometer records movement. The footage is subsequently analysed using AI vision models capable of identifying potholes and assessing them by size.The technology is particularly useful for detecting road defects that may not be immediately obvious to the human eye. Some potholes, when viewed from a distance, can even resemble speed breakers, making them difficult to distinguish during a moving commute.
AI spots the pothole, but Sen wanted the next answer
Detecting damaged roads, however, was only the beginning. The bigger challenge was one that commuters encounter repeatedly after filing complaints: Who is actually responsible for repairing this road?To address that question, Sen modified the application to work through around 2,900 government contracts. The system can use the location of a pothole to identify the contractor responsible for the particular stretch of road.That additional layer could be important when a road is still covered by a contractual warranty. In such cases, the contractor may be required to rectify defects without the government having to incur another expenditure.The technology therefore attempts to move beyond simply documenting bad roads. It connects a visible defect with the paperwork that determines responsibility for it.
A pothole can become a four-second data record
Sen’s application is designed to turn the information collected during a drive into a structured record in about four seconds.The record can include the pothole’s photograph, its precise location, the relevant tender number, and the officer responsible for the road.That changes the nature of a civic complaint.Instead of a resident reporting that a particular road has been damaged, authorities could potentially receive a complaint supported by visual evidence, geographical coordinates, and contract-related information.The difference may appear small, but it can be significant. Finding a pothole is one task; establishing who has the obligation to address it is another.
One commute produced 12 documented potholes
The system was eventually put through a real-world test when Sen used it during his drive to work.It detected 12 potholes along the route, generating documented evidence that could subsequently be used while approaching civic authorities.The demonstration offered a glimpse of how AI could change the way everyday urban problems are reported. Rather than relying entirely on residents to manually identify, photograph, and document road damage, technology could potentially perform much of the detection and documentation automatically.
From Bengaluru’s roads to a wider civic-tech idea
Sen demonstrated the system in a video shared on Instagram in collaboration with ChatGPT India. The video later reached X, where it attracted considerable attention.The response also showed why the idea resonated beyond Bengaluru. Potholes may be a local problem when encountered on a particular street, but the difficulty of identifying responsibility for damaged public infrastructure is much broader.Some users suggested that civic authorities themselves could deploy similar technology to survey roads, identify damaged stretches, and potentially accelerate maintenance work.That possibility could make the experiment more consequential. A system that detects potholes automatically could, in theory, help create a continuously updated picture of road conditions rather than depending solely on sporadic complaints from citizens.
When AI turns a road hazard into a paper trail
There is, of course, a long distance between detecting a pothole and repairing it. Technology cannot substitute for road maintenance, administrative action, or contractual enforcement.But Sen’s experiment addresses a crucial part of the process: information. A pothole that is photographed, precisely located, and linked to a tender and contractor is harder to treat as an anonymous defect. It becomes a specific problem attached to specific records and, potentially, specific responsibilities.What started as an attempt to make one engineer’s daily commute a little smoother has consequently become something larger, an experiment in using AI to make civic complaints more precise, more evidence-driven, and potentially much harder to ignore.
