An IoT system to curb vandalism on transmission networks: A case of UETCL

Abstract

Utility companies in Uganda are becoming increasingly concerned about vandalism of their assets. The Uganda Electricity Transmission Company Limited (UETCL) faces vandalism of high-voltage transmission towers, which frequently results in equipment damage, high maintenance costs, and service disruptions. An Internet of Things (IoT)-based system for real-time detection and alert of possible vandalism attempts on transmission infrastructure is designed and developed as part of this project to mitigate vandalism. The system comprises various sensors for detection of different parameters around tower infrastructure. A PIR motion sensor for motion around the tower, an accelerometer for tilt and vibrations on the system case and vibration sensors for vibrations due to cutting or banging on the tower. The system also employs a camera for visual observation. A central processing unit, the Raspberry Pi 5 is used. It accepts sensor and camera inputs, processes data and sends output signals to allocated recipients. This Pi also hosts a machine learning module which differentiates between human invaders and other stray objects. Tower status updates are transferred using MQTT protocol over cellular network, to a specially designed web dashboard, where personnel can observe real-time alerts, photographs of on-ground tower status and historical data for analysis and establishment of vandalism patterns on the network. Sensitive system components including the Pi, battery and other components are housed within a tamper proof casing to prevent damage. The system is powered either by solar or through induction powering from high voltage lines. In either scenario, a battery backup is provided in case the power source goes off, thereby ensuring uninterrupted system function. During this project, system requirements were determined, a modular architecture was designed, the system was developed and tested in various settings, simulating vandalism scenarios around Uganda’s transmission towers. Results obtained indicate the system’s capacity to reduce transmission tower vandalism through efficient real-time detection, dependable communication, and precise classification of the tower environment by the machine learning module, which collectively enable stakeholders devise faster and more efficient methods of response to vandalism attempts. Future work will focus on system enhancement in