AI- Powered Pothole Repair: A Sensor-Driven Approach to Infrastructure Maintenance

Potholes pose severe threat to road safety and maintenance, especially in developing nations where manual repair methods are slow, labour-intensive, and inconsistent. The design and development of a semi-autonomous pothole filling robot that can locate, estimate, and fill potholes with minimal assistance from humans is presented. The robot combines a mobile base with DC motor-controlled wheels, camera-based vision system for sensing, and a gantry mechanism for precise deposition of materials. According to pothole area and depth estimates obtained via image processing and ultrasonic sensing, a gantry-mounted hopper with a volume-controlled extrusion valve releases the patching material into the pothole. Compaction is achieved through a pneumatic actuator system. The robot has a dual microcontroller architecture involving Raspberry Pi5 for image processing tasks and Arduino Uno for navigation tasks. The robot demonstrates effective semi-autonomous performance for small to medium-sized potholes. The system provides a cost-efficient and scalable way of automated road maintenance and opens the door to future developments in autonomous repair robots.