Developing a Drone-Based Network Infrastructure Inspection System for Telecommunications Towers in North East Nigeria

Telecommunications tower inspection in remote and security-challenged terrain in North East Nigeria is infrequent and hazardous for human tower climbers, resulting in delayed fault identification and prolonged network outages. This study developed a drone-based automated infrastructure inspection system for telecommunications towers in North East Nigeria, targeting the Borno and Yobe State tower fleets. A professional UAV systems engineering methodology was applied, selecting a DJI Matrice 300 RTK as the inspection platform equipped with a Zenmuse Z30 30x optical zoom camera and a Zenmuse XT2 thermal camera. An automated flight path planning algorithm was developed in Python using the DJI PSDK API, generating pre-programmed inspection patterns around tower structures based on tower height, leg configuration, and antenna count metadata from the operator's asset database. A computer vision inspection pipeline using a YOLOv5 model fine-tuned on 1,800 tower component images detected structural anomalies including corroded bolts, damaged antenna mounts, and fraying transmission lines. The inspection pipeline achieved 88.7 percent detection precision and 84.3 percent recall on a held-out validation set. A full tower inspection cycle (pre-flight, inspection, data download, and report generation) was completed in 34 minutes, compared to 4 to 6 hours for a manual climbing inspection. Report generation produced a structured defect log with severity classification and photographic evidence. The study recommends the Nigeria Communications Commission explore drone inspection standards as part of tower safety regulations and facilitate NCAA airspace authorisation procedures for routine telecom tower UAV inspections.

Keywords: drone inspection, telecommunications tower, computer vision, North East Nigeria, UAV infrastructure