Design and Implementation of a Semi-Autonomous Robotic System for Systematic Solar Panel Cleaning

Solar photovoltaic (PV) systems have become a trend as a sustainable energy source on a global scale. However, the dirt accumulated on solar panels often reduces their efficiency, considerably. Existing cleaning methods have limitations such as requiring manual labour, low efficiency, limited to single cleaning method or single operational mode. As a solution, this research study describes the design and implementation of a semi-autonomous solar panel cleaning robot for both dry and wet cleaning. The robot can autonomously navigate on solar panels safely, or user can manually control it using a mobile app. The robot is capable of functioning for around 75 minutes on a single charge on solar panels tilted up to 20°, facilitated by image processing algorithms achieving an accuracy surpassing 96%. The orientation estimation is accomplished through the implementation of three parallel Kalman filters running on an embedded system. Experiments and simulations conducted using the implemented mobile robot show the effectiveness of the proposed robot design.