New Delhi: Deeptech firm drivebuddyAI has secured a patent for its Integrated Dynamic Road Quality Assessment System and Method, a technology designed to detect and geo-map potholes, rough patches and hazardous road conditions in real time, the company said on Sunday.
The Ahmedabad-based firm said AI-driven system combines sensor intelligence and deep learning-based computer vision to create a continuously updated road quality dataset tailored to Indian road conditions.
According to the company, the solution uses two parallel data streams. A GNSS sensor tracks vehicle speed and precise geo-location, while an IMU accelerometer captures acceleration changes across X, Y and Z axes. When a vehicle encounters a pothole or deteriorated road patch, the resulting Z-axis acceleration flags a road quality anomaly instantly.
To minimise false detections, the platform further processes video data linked to each anomaly through a fine-tuned deep learning model, visually confirming the defect and generating geo-tagged verified road intelligence. The company said the system continuously updates geo-locations as fresh data flows in, making the road quality map dynamic rather than a one-time survey.
“We are building AI that solves multiple problems at once. Road quality is not a standalone problem; it sits at the intersection of driver safety, cargo protection, and fleet efficiency,” said Nisarg Pandya, Founder and CEO, drivebuddyAI.
The company said conventional navigation platforms are optimised mainly for shortest time, based on distance and traffic, but do not account for road surface quality.
drivebuddyAI said its patented road quality intelligence layer will enable dispatchers and fleet managers to integrate verified road condition data into route planning, selecting routes that are safer and operationally faster in practice.
The company said the latest patent strengthens its portfolio of over 15 patents across ADAS-DMS and AI perception systems, including driver recognition and real-time drowsiness detection technologies.
