Abstract
Keywords
Introduction
Solar energy is a plentiful and dependable source of power, but atmospheric circumstances prevent it from operating at its maximum capacity. Solar panels, on the other hand, offer a way to capture and convert solar energy into useful electricity. Photovoltaic technology has emerged as a practical option to address the rising need for energy, among the many techniques. 1 Temperature, moisture content, and dust concentration are examples of weather-related parameters that have an impact on solar panel efficiency.
The performance of photovoltaic panels is greatly impacted by dust collection. 2 Research into the relative merits of solar panels with and without dust collection ultimately led to the creation of a mechanism that automatically cleans the panels. These problems have led to the development of an automated system for cleaning solar panels. An automated system for cleaning solar panels ensures thorough cleaning, eliminating performance issues caused by dust buildup. Solar panel performance under varying dust collection conditions (daily, weekly, monthly, etc.) has been the subject of research. Keeping the panels clean greatly improved their efficiency, as seen by comparing their performance before and after cleaning. Cleaning technologies that utilize the Internet of Things are transforming the maintenance and operation of solar panels. Manual, time-consuming, and sometimes inefficient traditional cleaning methods are still used today. 3 This state-of-the-art system maximizes energy production and decreases maintenance costs by automating and optimizing the cleaning process with the use of Internet of Things (IoT) technologies, sensors, and connections. Data on dust accumulation, solar radiation levels, and weather conditions are just a few of the variables that the IOT-based solar array cleaning system monitors in real time through a network of attached sensors. 4
The data analysis shows that these systems activate, enabling targeted cleaning. 5 Operators and maintenance workers can receive data and status updates in real time using the remote monitoring and control functions of the IoT-based system. 6 The user can receive notifications or warnings to remotely oversee the cleaning process and make any necessary adjustments. As a result, proactive maintenance can handle any problems quickly, and the system as a whole is more efficient. An IoT-based array cleaning system tracks and reports in detail the performance of solar panels, cleaning cycles, and energy output. This data enables cleaning schedule optimization, pattern identification, and data-informed system maintenance and improvement decisions. 7 When dust settles on solar panels, it can drastically reduce their efficiency and effectiveness. Dust covering solar panels blocks some of the sunlight that would otherwise reach the photovoltaic cells, reducing their efficiency. 8 When there is dust on the surface of solar panels, the photovoltaic cells have a harder time efficiently turning sunlight into power. A dust coating reduces the panels’ overall efficiency by acting as an insulator and preventing them from absorbing light as efficiently. Localized shadowing could occur if dust were to settle unevenly on the solar panels’ surfaces. When the panel’s electrical output is uneven due to shading, hotspots might emerge, potentially damaging cells. 9
Literature review
It is possible to increase the efficiency of each solar panel, which would reduce the total number of panels needed and lead to substantial savings. Not only does this approach improve efficiency, but it also helps planet. 10 Maximizing the output of solar panels requires attention to two critical factors. Getting as much direct sunlight as possible onto the panels should be top priority. Optimal placement, orientation, and tracking devices are required to ensure that the panels absorb as much sunlight as possible during the day. Importantly, the received light energy must be efficiently transformed into electricity. 11 For this, we will need top-notch solar cells and a power conversion system that can keep energy losses to a minimum.
To maximize the efficiency of solar panels, one can clean them using various methods. These methods can be impacted by the amount of dust or grime. 12 As part of the cleaning process, water is used to wet and rinse the solar panels The method can range from utilizing a hose to more complex tools like automatic sprinklers or water jets. Use a soft-bristled brush or broom to clean solar panels by hand; be careful not to harm the surface. To remove dirt and debris from solar panels, automated cleaning systems use mechanical parts that glide across the surface, including wipers or brushes that spin. The use of hydrophobic coatings is another strategy for keeping solar panels clean. The panels remain clean because dust and debris are unable to adhere to the panels’ non-stick surfaces, which also repel water. When choosing a method to clean solar panels, it is crucial to take the manufacturer’s instructions into account. The type of panel, the terms of the warranty, and the surrounding environment are all important considerations. 13
Fitriana et al. 14 developed a prototype of a wiper cleaner for cleaning solar panels. The ESP8266 limits the range of WiFi connectivity. The use of the Blynk application is a significant element affecting this connectivity range. Additionally, using the wiper to clean solar panels might leave surface scratches, especially when there is a lot of dust present.
Khadka et al. 15 worked on a system that rolls over the edges of PV panels. A microprocessor controls the robotic unit’s prototype. The internal Wi-Fi antenna connects to a wireless network and can only connect to the cloud. Their effort involved several extensive tests carried out on a 50 W PV panel. The researchers used a regression model to analyze the results. An important finding from the investigation was that PV panels that remained dust-free produced significantly more electricity than those that had dust buildup. 16 The study also revealed a significant difference in power production between the two scenarios in cloudy weather. However, the power production difference significantly decreased when the panels were exposed to bright weather.
Anilkumar et al. 17 collaborated on a study that led to the creation of a solar panel robotic arrangement-specific wireless networking system. Anilkumar et al. meticulously partitioned and controlled each part of the system with its own NodeMCU unit. These individual NodeMCUs linked to a central NodeMCU served as an orchestrator or hub for assembly. 18 The web-based program allowed users to command the robot’s movements through a human-machine interface. With this web-based HMI, users could control the robot’s functions from a distance. Using a web-based approach would increase accessibility and convenience since anyone with a regular web browser could simply interact with the system. Pay close attention to the limitations that come with employing a web server-oriented application. 19 Importantly, the maximum functioning range of the apparatus was a pitiful 10–15 m. As stated in the range constraint, the controlling device, or the central NodeMCU, could only command the robot to its full potential within this extremely narrow range. 20 This study and its accompanying system design shed light on a novel application of robotics and wireless networking to the solar panel manufacturing process.
Kassim and Lazim 21 worked on a web-based monitoring system. The effects of shading, sunlight absorption, and the amount of power it generates reduce a static solar panel’s effectiveness. They intended to design and connect the automatic adaptive PV module prototype and an efficient algorithm tracking system via IOT. 22 The system has accomplished its focus on efficiency optimization. A tracking system that uses an active technique orientation enables the capture of more power and energy. An Arduino microcontroller was monitoring the high solar panel voltage, and a light-dependent resistor was measuring the voltage in line with the solar rotation angle. 23
The solar panel cleaning method described in the research 24 offers a novel approach to ensuring that commercial and industrial-scale solar installations continue to provide the maximum amount of power possible. The device’s power source is a rechargeable battery, and a mobile app can easily activate it for convenient remote control. The mobile app accurately controls the cleaning tool’s horizontal movement by sending output signals to the gear motors through an internet connection. 25 Rooftops, carports, and other elevated constructions sometimes have restricted access to water and power; thus, this is very helpful in those situations. A linear piezoelectric actuator is integrated into the system to tackle the unique problems of these situations. 26 It cleans the system efficiently and allows the solar panels to work at their best regardless of the weather. This complex method efficiently controls the flow of energy while also optimizing the solar power source, the load, and the hybrid storage system that combines supercapacitors and batteries. The end product is an all-inclusive strategy to reduce dirt buildup on the panel surface. Solar panels can lose some of their efficiency as dust, filth, and other pollutants build up; the cleaning system is a practical solution to this problem. To clean the solar panels properly without damaging them, use water and a gentle, nonabrasive sponge or cloth. This will get the job done well without causing any harm. 27
Designed specifically for off-grid solar PV systems, article 28 details the architecture of a cost-effective remote monitoring system. To create the remote monitoring system, researchers set up a network of sensors on a router that meets the IEC-61,724 standards, along with an Arduino Uno Wi-Fi communication system for a demilitarized zone. It can measure electrical and weather conditions with an error rate of 2% or less. The suggested monitoring system primarily consists of four components. The first step is to configure the router so that the Arduino Uno Wi-Fi can communicate with a DMZ. Additionally, it ensures that the average error stays below the 4% threshold by meeting the minimum current, voltage, power, and PV-module temperature parameters specified by the IEC-61,724 standard. Thirdly, it guarantees a degree of precision for electrical and meteorological parameters of 98%. When compared to commercial data logging equipment, this one is far more accurate. New processing, display, and access strategies, as well as the option to choose the number of measurements handled per hour, are some of the highlights. 29 With these additions, data stored on campus computers and mobile phones can be accessed for pedagogical and scientific reasons. 30
Lightwala et al. 31 introduced an automated dust detection and cleaning system with the potential to clean PV modules. Since dust hinders solar energy from reaching the surface of the module, reducing system performance, it is one of the variables that negatively impact PV module output. 32 The authors built the system using C software, which we designed and compiled with Arduino IDE. 33 This program reads voltage, senses current from the PV, and calculates power output. We employed Proteus 8 Professional to construct the circuit. When dust accumulation on module surfaces caused power loss, the system activated the motor drive of the cleaning mechanism.
Jaswanth Yerramsetti et al. 34 suggest methods for cleaning solar panels without human involvement. This method makes it easier to clean floating aquatic plants, where a human can’t clean them. The robotic car uses internet communication to function as a manually operated vehicle. Additionally, this robot is automatable. 35 The gear motors and motor driver power the robot, and a separate motor connects to a cleaning brush to wash it with water. The robot takes the water from the body of water. Additionally, the Raspberry Pi can be used to regulate whether cleaning is done with or without water and also regulate the system’s operation time. 36
This study 37 presents a new and effective solar cleaning robot that uses a four-side stretch sling. Its purpose is to address the long-standing problem of how to keep solar power systems operating at peak efficiency. The goal of the robot’s design was to make large-scale solar panel installations easier to clean by cutting down on human labor, operating expenses, and configuration flexibility. As it glides across solar panels, the robot’s water sprinkler system and spiral roller brushes guarantee thorough cleaning coverage. 38 The novel four-side stretch sling system allows for precise movement and accurate localization, which contributes to its usefulness in different contexts. It features a length measurement unit and in-built control systems. 39 The robot showed an accuracy level of 95%–99% in tests that included controlled releasing and pulling procedures and manual control with a radio frequency remote. Solar farms, rooftop solar panels, and solar floating panels are just a few examples of the many potential uses for this versatile robot. In sum, the findings of this study provide encouraging evidence of progress toward improving solar power systems’ efficiency and feasibility via novel robotic solutions.
This article presents a fully automated IoT-based solar array cleaning system that utilizes the Messaging Queuing Telemetry Transport protocol to enhance the efficiency of solar panels. The Adafruit dashboard effectively constructed and managed the system. The system response was prompt, with a minimal delay of 2 s. The implemented method possessed the capacity to greatly enhance the efficacy of solar panels. Throughout the testing process, we observed that cleaning the solar panels increased the power output compared to panels covered in dust. The use of IoT technology enabled the continuous monitoring of system performance in real time. Operators oversee and regulate the cleaning activities remotely, minimizing the necessity for in-person visits. The use of this technique resulted in a 30% enhancement in the solar output of a 30 W solar panel.
Further contents of the paper are organized as follows: methodology is described in Section 3, system design and hardware are discussed in Section 4, results and discussions are explained in Section 5, comparative study in Section 6, and the conclusion of the paper is described in the last section.
Methodology
Figure 1 illustrates the detailed architecture of the advanced cleaning system. This block diagram depicts a consolidated cloud-based platform that processes and stores data from the Internet of Things system. Its handling of data storage, processing, and display enables remote monitoring and control. A variety of sensors, including those for dust, voltage, and precipitation, make up the sensor network.
Block diagram of system.
The ESP-32 microcontroller is the pivotal component of the system. The ESP-32 retrieves data from sensors and manages various system activities. The system uses the rain sensor FC-37 to save energy and resources. During wet seasons, the system can refrain from active cleaning. The rain sensor keeps tracking the weather even after it stops raining. The GP2Y101 dust sensor gives a picture of how clean the panels are and sets off cleaning cycles when dust levels go too high. The system initiates the cleaning procedure mechanically the moment the dust sensor identifies a substantial accumulation of dust. The voltage sensor model F031 is essential to the system because it keeps tabs on the voltages of the solar panels and the batteries. By linking to the battery terminals and taking constant voltage readings, the sensor enables the system to estimate the battery’s charge level and health.
The water pump is used to sprinkle water over PV panels. The amount of water sprinkled is controlled by the dust sensor. Water is only used for cleaning when there is enough dust. Schedules, sensor readings, or real-time data prompt the water pump to initiate cleaning cycles. Setting the system to activate the pump only when needed achieves water savings. An LCD and an I2C module integrate into the system to facilitate monitoring and controlling the parameters. The LCD connects to the microprocessor and displays current data, including the voltage state of the system, the voltage of the solar panels, and the battery. The I2C module streamlines connectivity and improves efficiency by simplifying communication between the microcontroller and the LCD.
The system utilizes gear motors to rotate photovoltaic panels using mounted wheels. Gear motors enable the system to achieve bidirectional motor operation. If there is a significant accumulation of dust on the surface of the PV panel, the system will operate at a reduced speed to perform a thorough cleaning. To carry out the cleaning process, the DC motor connects to soft nylon brushes. These motors rotate the brushes to carry out the cleaning process. The quantity of dust present regulates the speed of the cleaning brushes. A 10W solar panel, along with a 12 V battery and charge controller, powers the system. A charge controller is a device that regulates the flow of power and manages the charging process of a battery. The system does not immediately function with a 10W solar panel. A 10 W panel alone charges the battery, which subsequently supplies power to the system.
Figure 2 illustrates the system flowchart, providing a detailed explanation of the system’s operations. Upon booting up, the ESP-32 microcontroller initiates a WiFi search and subsequently establishes a connection with the MQTT. Once the ESP-32 obtains the sensor readings, it proceeds to transmit the data to the MQTT data streams. When MQTT receives subscriber queries, it verifies the operating mode of the system. If the system is not in sleep mode, it initiates the cleaning process.
System flowchart.
Adafruit IO is an IoT project-specific cloud platform. Its straightforward and user-friendly interface makes it easier to connect and operate IoT devices, gather data from sensors, and create interactive dashboards and apps. Popular development boards like Arduino, Raspberry Pi, and ESP8266/ESP32 are among the many IoT devices and platforms that Adafruit IO supports. For makers, amateurs, and IoT enthusiasts, it streamlines data collection and visualization from your devices.
MQ Telemetry Transport, often known as MQTT, enables the transfer of telemetry data even in situations with limited bandwidth. Network users can depend on MQTT, a lightweight and open messaging protocol, for a reliable and efficient communication pathway with minimal resources. The system utilizes the subscriber request to obtain the command signal and employs the publisher technique to transmit the data to the storage.
Sleep and active modes are the high-power and low-power modes of the proposed system. In low-power mode, only basic components will be in operation, while in active mode, the system will use its full potential. The primary objective of the IoT system is to enhance the efficiency of solar power generation. However, it is imperative to consider its inherent environmental impacts. The design and methods used in the IoT-based system can constrain the effectiveness of the cleaning process. Some dirt, stains, or debris may require manual cleanup.
System design and hardware
Multi-layer circuit design involves creating electrical circuits with numerous layers of insulating and conducting lines. Frequently used in intricate electronic systems with limited space and a high density of circuits, compared to single- or double-layer circuits, multi-layer circuits enhance size, signal integrity, and functionality.
Figure 3 shows the three layers of circuits connected in parallel and spaced apart by copper braces. The microcontroller links every sensor.
Multi-layer circuit.
Figure 4 shows the first layer of a system, which consists of the following electronic components:
Development Board ESP-32 Wroom.
LCD 20 × 4.
I2C Module
Voltage Sensor
Top layer circuit.
These parts work as the system’s primary layer, providing features like control of the microcontroller, display output, communication abilities, current sensing, and voltage sensing.
Figure 5 represents the second layer of the system, encompassing the following components:
Voltage Sensor
Buck Converter
Boost Converter
Dual Relay Module
Second layer of circuit.
These parts work together to provide the second layer of the system, which enables features like voltage detection, voltage conversion (both step-up and step-down), and control over the electrical circuit via the relay module.
An efficient method of controlling the power supply is by using a buck converter. The main goal is to stabilize the voltage at 5 V from the 12 V output of the battery. The sensors and the ESP-32 controller, among other system components, rely on this voltage drop to function. The ESP-32 functions more reliably and efficiently within its voltage range because of the steady 5 V supply it receives. The boost converter steps up the voltage from 5 to 12 V, providing the necessary power to run motors and motor drivers. To manage the system’s power modes, a relay module is essential since it allows for the changeover from a low-power sleep mode to a high-power active mode. It is impossible to maximize energy usage and overall system efficiency without this feature.
Following Figure 6 is the third layer of the system, which includes the following components:
Motor Drivers L298N
Connectors for Motors
Third layer circuit.
These components form the third layer of the system, enabling the control and operation of multiple DC motors through the L298N motor drivers.
Figure 7 displays the electronic component box of the system, providing an overview of the crucial parts that impact its functionality. It is worth mentioning that the power supply houses a battery behind the charge controller. An intentional decision to reduce the system’s total weight was to build the electronics box out of acrylic glass.
Electronic box of the system.
Figure 8 shows a top-down perspective of the system, depicting a 10 W solar panel, a charge controller, and a three-layer circuit board. To better comprehend the system’s architecture and configuration, this graphic depicts its spatial structure and the connections between its components.
Top view of the system.
Figure 9 shows the project’s final hardware. The solar panel is installed on top of the metal body, providing self-power to the system. An acrylic sheet circuit box is attached to house the circuit, battery, and charge controller. Its self-powered mechanism is what makes the system innovative. A 10 W solar panel and a battery power the system. The primary characteristic of the system is its ability to manage a worldwide network through the utilization of Adafruit IO. Users can control the system from anywhere on the globe, as long as they have a device linked to the internet.
Final project hardware.
Results and discussions
Figure 10 depicts a dashboard for Adafruit in visual form. Users can control the cleaning system through buttons and indicators on the dashboard. The informative indications, which offer real-time feedback on the power condition of specific modules, are located on the left side of the dashboard. Notably, the prominent “System Power” indication, which serves as a visual cue for the system’s current state, dynamically changes its color. The indicator glows a vivid green color when it is in the active state. On the other hand, the indicator takes on a calm red tint when the machine goes into sleep mode.
Adafruit active buttons.
The Sleep/Active button regulates the energy usage of the system. During active mode, the device utilizes its maximum power to carry out the cleaning process. Sleep mode reduces power usage by powering off the system while keeping only the sensor and microcontroller active. The water pump button enables the user to alternate between the dry and wet cleaning processes. The system is capable of halting the cleaning brushes in the event of a system malfunction without requiring a complete shutdown of the entire system. This control enhances safety measures during system maintenance.
Figure 11 shows some visuals that users can use to keep tabs on and examine data from their connected devices. The line charts help users easily visualize how certain parameters change over time by placing data points on a graph. By placing data points on a graph, users may easily see how certain parameters change over time. For specific parameter values, the gages serve as real-time counters. The gages update their data every 10 s to bypass Adafruit IO’s data throttling limitations, which restrict access to 30 data points per minute. By updating the data every 10 s, customers can be confident that the data they receive from their linked devices will be up-to-date and accurate. The
Adafruit storage graphs.
Figure 12 shows a graphic depiction of the live data monitoring and account status capabilities of the Adafruit platform. A user-friendly dashboard or graphical user interface, providing users with real-time insights into their IoT device data and the ability to detect faults in the data stream, is a part of this display.
Adafruit account status & live data.
However, the Adafruit IO platform has certain limitations. The primary restriction is the maximum number of approved Internet of Things devices, which is 2. Additionally, users can only use a maximum of 10 data sources, and the data rate must not exceed 30 points per minute. As a result, we can only display 30 points’ worth of data in a minute. Users need to prepare ahead of time and distribute the devices and data flows properly to ensure the system operates well within these limits.
Figure 13 shows the ESP-32’s serial monitor, which demonstrates establishing a WiFi connection, utilizing MQTT, and transmitting data to the cloud. It communicates at a baud rate of 115,200. At initialization, the ESP-32 uses the provided SSID and password to try to connect to the selected WiFi network. Upon establishing a WiFi connection, the ESP-32 displays the assigned IP address. This IP address is required to connect to local and internet networks. To initiate message exchanges with the MQTT server, the ESP-32 requires internet connectivity. The MQTT protocol enables an efficient connection between the ESP-32 and the server.
Serial monitor sending data.
The ESP-32 collects data from the connected sensors once the WiFi and MQTT connections are established. After that, it sends the sensor data to the Adafruit IO channels. Throughout the data transmission procedure, the ESP-32 provides real-time status information on the value-sending operation. Thanks to this feedback, users may monitor the data transfer’s success or failure.
Figure 14 depicts the reception of a subscriber request from the Adafruit dashboard, coupled with the associated indication of execution. To fulfill a subscriber request in Adafruit IO, the user must register for a specific data feed. Subscribers receive immediate updates whenever new data is published to the feed. Users have the option to select a particular data stream that they wish to monitor and receive updates from. Adafruit IO monitors the selected data feed for any new data. Whenever a user activates the button on the dashboard, Adafruit IO monitors the selected data feed for any new data. The ESP-32 receives an instruction to either activate or deactivate a specific stream. The microcontroller thus oversees the system based on the subscribed data stream.
Serial monitor subscriber requests.
Our suggested system has a very low average delay of just 2 s, as shown in Table 1. Testing across several networks consistently proves the system’s resiliency, as the controller and the controlling device deeply connect to the same network, guaranteeing a smooth communication process. This negligible 2-s delay, which stands out when compared to competing approaches, emphasizes the effectiveness and responsiveness of our suggested configuration. Because of this small lag, the system works better in real time and is better suited for scenarios that call for coordinated control.
Data of delay test results.
The output voltage and current values for a 30W solar panel coated with a thick layer of dust and bird droppings are displayed in Table 2. According to the data, the panel’s output power has dropped significantly, highlighting how the debris has been affecting it. Multiple readings were collected at different periods to fully grasp the varied production under these conditions, while the solar panel dust condition remained nearly the same during each test.
Experimental data of solar panel output power with dust.
The data in Table 3 shows that cleaning a 30W solar panel results in increased light absorption and power production. The findings demonstrate that a clean solar panel may absorb more light and produce more power. The data clearly shows that keeping the solar panel clean increases its efficiency and electricity output.
Experimental data of solar panel output power without dust.
Comparison with existing works
A comparison of the three prior projects was done to identify and quantify the similarities and differences between our project and the others in terms of communication, water system, power-saving mode, control mode, microcontroller, cloud storage, control app, and system power. Our system offers both auto and manual modes, allowing remote control without the need to visit the site. Our system has the advantage of having two power sources: a battery and a solar panel, while other systems are only powered by rechargeable batteries. Adafruit cloud storage gives a large amount of memory to store and analyze the data, but other approaches, like the Blynk application, only provide access to store a few data points. Table 4 presents a comparative analysis.
Comparative analysis.
When comparing IoT-based solar array cleaning solutions to other conventional or automated approaches, one might observe a multifaceted landscape of advantages and disadvantages. The efficiency of IoT solutions is amazing. By utilizing real-time monitoring and data-driven decision-making, they can optimize cleaning schedules and energy generation. However, the increased productivity associated with IoT comes with a significant drawback: the initial investment required for integrating IoT devices and infrastructure is considerably higher compared to old manual methods or targeted automated solutions that do not involve IoT components. Increasingly, there are growing concerns regarding power consumption in IoT equipment.
Conclusion & future work
In this article, a fully automated IoT-based solar array cleaning system was proposed using the Messaging Queuing Telemetry Transport protocol to improve the efficiency of solar panels. The system was effectively constructed and managed using the Adafruit dashboard. The system response was prompt, with a minimal delay of 2 s. The implemented method possessed the capacity to greatly enhance the efficacy of solar panels. Throughout the testing process, it was seen that solar panels that had been cleaned produced a greater amount of power compared to panels that were covered in dust. The use of IoT technology enabled the continuous monitoring of system performance in real time. Operators oversee and regulate the cleaning activities remotely, minimizing the necessity for in-person visits. The use of this technique resulted in a 30% enhancement in the solar output of a 30W solar panel.
The future of the solar array cleaning system, which is based on IoT technology, shows great promise and has the possibility for further enhancement. Researchers can prioritize the improvement of cleaning mechanisms by exploring novel materials and cleaning agents. By monitoring performance and weather data, the integration of predictive maintenance capabilities can proactively identify and anticipate issues before they occur. With the increasing number of solar installations, it is important to design scalable solutions that can be easily integrated into current systems. This will result in a more streamlined and flexible cleaning system, promoting the development of renewable energy solutions.
Footnotes
Declaration of conflicting interests
Funding
ORCID iD
