From pv magazine International

Researchers from the University of Johannesburg in South Africa have developed a cleaning system for solar modules based on the TCS3200 color sensor and the Arduino Uno open-source microcontroller.

The TCS3200 is a programmable color light-to-frequency converter equipped with white light-emitting diodes (LEDs) to illuminate an object’s surface for color detection, with the object reflecting light to the sensor to determine color intensities.

“TCS3200 can detect the color of the light incident and output square waves with 50% duty cycles,” the scientists said. “Compared to other scientific instruments, this sensor is reasonably at a low cost.”

The microcontroller is the key element of the entire system, which also includes a 46 V motor driver, two 12 V DC motors, a charge controller, a rechargeable battery, a voltage regulator, and a roller brush.

“The roller brush provides a sweeping motion for cleaning the panel surfaces by spinning at a higher speed than the movement wheels,” said the research team.

The system also uses Arduino IDE software to write, compile and upload sketch codes for the system. It includes the connecting rods, the case shielding the electronic components, and the brush-driving devices.

The system can detect dirt on solar panels by measuring colors, calibrating colors, and comparing measured results with reference results.

“When the controller detects that the humidity threshold level signal has reached 80%, the motor driver will turn on to clean the solar panel,” the academics explained. “Motor A of the two DC motors in the setup moves the color sensor across the solar panel for color monitoring, while Motor B drives the cleaning robot to remove dirt.”

The group estimates annual system costs at €1.50 ($1.60) per panel, with capital and maintenance costs accounting for most of the associated costs.

“In one minute, up to 95% of the dust on the PV panel surfaces can be removed by this device,” claimed the group.

The scientists described the new cleaning system in “Solar panel surface dirt detection and removal based on Arduino color recognition,” which was recently published in MethodsX.

“The researched project is an improvement over existing systems since it eliminates many of the drawbacks, such as the need for water, manual cleaning, the need for labor,” the scientists concluded. “The system can operate for many years before requiring maintenance.”