Abstract: To address the posture and motion adjustment challenges of wall-climbing inspection robots during Phased Array Ultrasonic Testing (PAUT) and Time of Flight Diffraction (TOFD) for welds on the top and side walls of large pressure-bearing equipment, this paper designs and develops an intelligent weld-tarcking automatic wall-climbing inspection robot based on contour sensors, intergrating laser weld tracking technology. Based on the extraction of detection results from the laser ccontour sensor, a multi-scale linear fitting method is applied to capture mutation points in the data sequence, enabling rapid weld identification. The relationship between the laser weld tracking data and the linear velocity of the mechanical drive module of the crawler is analyzed to realize the crawler’s intelligent tracking of the weld path. Test results demonstrated that the developed inspection robot can follow both straight and curved welds during movement. Moreover, the positional deviation of the robot’s intelligent weld tracking meets the requirements for industrial inspection, ensuring complete acoustic beam coverage of the weld area and guaranteeing the validity of the weld inspection data.