Bengaluru engineer Gaurav Sen has built an AI tool that detects potholes through dashcam footage during daily drives. He demonstrated the system on Instagram on August 16, 2026, to tackle Bengaluru’s road problems using AI, GPS and vehicle sensors.
The system also finds contractors responsible for damaged roads by checking government contracts.
Sen built the app with OpenAI Codex and connected it with vision models to study road footage. A dashcam, GPS and accelerometer collect data while the car moves through Bengaluru streets. The AI can spot potholes that may look like speed breakers from a distance.
The system goes beyond simply finding damaged roads and also checks who should repair them. Sen added data from 2,900 government contracts to identify contractors linked with specific road stretches.
This feature could help with roads still covered under warranty, where contractors may need to fix defects. The system can create a detailed record within four seconds after finding a pothole.
Each record includes a photograph, exact location, tender number and responsible road officer. Sen tested the tool during a drive to work and found 12 potholes along his route.
The results could provide stronger evidence for complaints sent to civic authorities. The project also gained attention after the video reached X and other social media platforms.
A social media reaction described the project as ‘Fixing real-life bugs.’ Another user wrote, “Great idea and execution.” Users also hoped the system could help improve road repairs across Bengaluru and other cities.
Sen’s project gives a practical use to AI beyond regular apps and chatbots. It connects road damage with location data and contractor details, creating a simpler way to report potholes.
Also Read: Best 4K Dashcam Deals for Safe Driving: Amazon Weekend Offers