Abstract
The increasing age of petroleum pipeline infrastructure poses significant risks to safety, operational efficiency, and the environment. Defects such as cracks, corrosion, joint displacement, and deformation remain major causes of leakage and catastrophic failure. Traditional inspection approaches rely heavily on manually operated robotic crawlers, with defect detection dependent on human review of video footage, resulting in time-consuming and error-prone assessments. This paper presents the design, fabrication, and experimental validation of an autonomous modular crawler robot integrating a Raspberry Pi 4, Arduino Mega, high-resolution camera, ultrasonic distance sensors, and gas detection capabilities for real-time, multi-modal defect detection in petroleum pipelines. The proposed system achieves autonomous navigation, real-time video streaming, and multi-sensor data fusion, enabling robust inspection in varying pipe diameters and material conditions. Laboratory and simulated field experiments demonstrated a maximum speed of 0.25 m/s, obstacle detection accuracy of 91.2%, climb capability of up to 45°, and battery endurance of approximately 80 min. Compared to existing inspection systems, the proposed crawler robot offers improved adaptability, sensing integration, and autonomy. The results position the system as a viable solution for preventive pipeline maintenance, with potential extensions into AI-driven defect classification and SLAM-based navigation.