Development of autonomous unmanned surface vehicle for online monitoring and evaluation of water parameters using GIS

In this thesis, an Unmanned Surface Vehicle (USV) system has been developed to enable real-time and autonomous monitoring of physical water parameters in aquatic environments such as seas and lakes. The system comprises multiple sensors capable of measuring key water quality parameters including temperature, pH, conductivity, dissolved oxygen and oxidation-reduction potential as well as components such as GPS for location tracking, LiDAR for obstacle detection, a digital compass for directional control, communication modules, a battery, and a power system. The software infrastructure has been developed using Python on the Raspberry Pi for the USV, C on the Atmega 324P microcontroller for the water measurement unit, C++ on the Arduino Uno for the control unit, and C#, JavaScript, HTML, and CSS for the web-based control center. Tasks such as motor control, obstacle avoidance, data transmission, and mission execution are performed using Python on the USV. The USV is designed to operate in both manual and autonomous modes and is managed through a web-based interface integrated into the control center. Through the map-based interface, the user can define mission points; the system then navigates to these points, conducts the necessary measurements, and transmits the data to the control panel in real time. Data transmission is achieved through communication technologies such as ZigBee and the GSM module at the hardware level, and SignalR and RESTful API at the application level. Additionally, the system includes a sample collection mechanism suitable for testing in laboratory conditions. The collected water quality data are processed within the scope of Geographic Information Systems using various interpolation methods, including IDW, Kriging, Spline, and Natural Neighbor, to generate real-time spatial distribution maps. This capability allows for the spatial analysis of environmental changes. In four different field tests, the USV travelled a total distance of 2,237.26 meters. When compared to the straight-line (as-the-crow-flies) distance, it was observed that the USV followed an optimal route with an accuracy of approximately 96.5%. The missions were completed in approximately 79 minutes, of which 44 minutes were dedicated to measuring water parameters and collecting samples. Out of 24 designated task points, the vehicle successfully reached 66% of them within a 3-meter tolerance. In obstacle detection tests, 6 out of 10 obstacles encountered were avoided without collision; in autonomous driving mode, all obstacles were avoided. A total of 120 different water parameter values obtained during field tests were processed in real-time through the control panel, and spatial distribution maps were generated using various interpolation methods such as IDW, Kriging, Spline, and Natural Neighbor. In addition, the parameter values were displayed in real time on the map. This thesis contributes a novel approach to environmental monitoring systems through a low-cost, modular, and user-friendly design, offering an effective and sustainable solution for both academic research and practical field applications.