In the context of the continuous development of green and low-carbon factories, large-capacity distributed photovoltaic panels and exhaust gas combustion treatment devices are often installed on factory rooftops. These installations significantly reduce external power demand, optimize the system’s energy structure, and alleviate environmental pressure. However, while generating economic benefits, they also pose certain threats to factory safety. The TFT-LCD (Thin Film Transistor Liquid Crystal Display) industry is characterized by large factory building areas, numerous equipment and pipelines, and complex operating conditions. Apart from large-area photovoltaic panels, the factory’s own exhaust gas treatment devices and air conditioning cooling units are also installed on the rooftop, imposing high safety requirements.
Photovoltaic panels on factory rooftops are prone to hot-spot effects, combiner box failures, and even fires due to the aging of some components or cables. Traditional manual inspection methods require high manpower and have low inspection efficiency, which cannot meet the increasing safety needs of factory rooftops. There is an urgent need for an efficient and feasible intelligent inspection method. This paper presents a comprehensive application of a smart inspection platform based on UAV drones, focusing on the long lifecycle management of rooftop equipment and proposing methods for visible light image analysis, infrared image analysis, and anomaly confirmation. The feasibility of the method was verified in a factory in Anhui, improving operational efficiency and yielding high economic and safety benefits.
The current status shows that the installation of large-capacity photovoltaic panels and exhaust gas combustion treatment devices may bring significant fire risks, including fire hazards caused by hot-spot effects of photovoltaic panels and abnormal electrical components of exhaust gas treatment devices. Common rooftop equipment monitoring methods mainly involve installing sensors at key parts of photovoltaic panels or exhaust gas combustion treatment devices. When data feedback indicates anomalies, maintenance personnel perform on-site inspections and temperature tests to confirm whether there is an anomaly. This manual inspection method has the advantage of being detailed and accurate in fault investigation. However, due to cost constraints, it still has disadvantages such as slow fault response, low detection efficiency, and high manpower demand. Moreover, because rooftop equipment often covers large areas and is high in height, manual inspections cannot cover all parts, leading to a small inspection range.
In the current context of increasing market competition pressure, not only does the cost of manpower increase, but there is also the hidden danger that equipment failures cannot be detected in time, leading to fires. For example, in April 2024, a fire broke out in the factory of an automotive electronics company in Zhejiang Province, with a burned area of about 210 m². The cause was a fault in the line under the photovoltaic panel that ignited combustible materials underneath. In November 2021, a sudden fire occurred in a battery factory in Shaanxi, caused by an electrical fault in the electric heating tape of the rooftop washing tower, igniting the tower body and surrounding combustibles.
Meanwhile, UAV drone technology is becoming increasingly mature and can serve as a new monitoring method in the field of factory safety. Compared with manual inspection, UAV drones have several advantages:
- High monitoring efficiency and low manpower demand: UAV drone inspection requires only one pilot to operate and cover the monitoring area, and the inspection efficiency is often more than 40 times that of manual inspection.
- Wide field of view: UAV drone inspection uses aerial photography for monitoring. Due to its high flight altitude, the field of view is much broader than manual inspection, covering areas that manual inspection cannot reach, such as the top of exhaust gas treatment device chimneys and the middle of photovoltaic panels.
- Infrared perspective and other functions not available to humans: UAV drones are equipped with dual-spectrum cameras and have functions such as distance measurement, effectively compensating for the shortcomings of manual inspection and achieving multi-perspective high-efficiency inspection.
This paper proposes a smart factory inspection method based on UAV drones, using multi-spectral cameras mounted on UAV drones for efficient data collection. The collected data is analyzed, combining infrared image comparison and visible light image comparison to determine the type of anomaly and guide electrical and exhaust gas treatment professional maintenance personnel to handle anomalies.
Establishment of Smart Inspection Method
Visible Light Image Analysis Method
UAV drones have a perspective advantage that manual inspection does not have. UAV drone inspection can set adaptive inspection routes based on the distribution of photovoltaic panels and equipment, considering parameters such as lens and flight altitude. Visible light images of equipment under normal operating conditions are collected by UAV drones to establish a visible light image database. Through the form of a safety checklist, a UAV drone inspection safety checklist is formed, as shown in Table 1.
By compiling the UAV drone inspection safety checklist, inspection personnel can effectively compare items one by one, significantly improving inspection quality and efficiency.
| Item | Content |
|---|---|
| General | No rust or damage on roof surface and pipelines |
| No abnormal smoke or fire on roof | |
| No abnormal personnel behavior on roof | |
| Photovoltaic Panel | No damage or cracks on the surface of photovoltaic panel |
| No shading on photovoltaic panel | |
| No weeds, seedlings or other debris on photovoltaic panel | |
| Exhaust Gas Treatment Device | No abnormal smoke from exhaust gas treatment device |
| Air Conditioning Device | No cooling water overflow from air conditioning device |
| No pipeline icing, etc. on air conditioning device |
Infrared Image Analysis Method
Based on the basic operating parameters and status of photovoltaic panels and exhaust gas treatment devices, the equipment operating temperature standards are formulated. According to the infrared image data collected by UAV drones, the local gray level analysis method is used to segment and extract hot spots from infrared images, outputting a list of abnormal temperature points.
The local gray level analysis method is described as follows:
From the principle of infrared imaging, the abnormal hot spots of photovoltaic panels and equipment that we are concerned about mainly appear as local high-temperature points against a low-temperature background. Therefore, based on the infrared image collected by UAV drones at a specified altitude, the image is divided into several local analysis units by pixels. Taking a local analysis unit, it is segmented and extracted according to the gray value, obtaining the gray abnormal range w1 within that local unit.
Since this local analysis method may produce gray abnormal points caused by the different temperatures of the edges of photovoltaic panels and equipment compared to the roof ground, which are noise points that need to be removed, to eliminate the influence of noise, as shown in equation (1), the gray abnormal ranges w1 to wn of several local analysis units are combined to obtain the combined gray range value W.
$$W = w_1 + w_2 + \cdots + w_n \tag{1}$$
Then, the combined gray range value W is compared with the area of a single photovoltaic panel or equipment unit W0:
$$N = W_0 – W \tag{2}$$
If N is negative, i.e., W is greater than W0, it indicates that this combined gray range is noise and should be removed, not participating in subsequent calculation and anomaly localization.
If N is positive, i.e., W is less than W0, it indicates that this combined gray range is an anomaly point, i.e., the hot spot to be paid attention to.
After detecting the hot spot region in the infrared image, according to the correspondence between the infrared image and the visible light image, combined with the visible light image, the defect type of the hot spot region is identified, including internal defects, cracks, or foreign object blocking, helping maintenance personnel quickly determine the repair plan, such as cleaning, repair, or replacement.
Long Lifecycle Monitoring and Management Method for Rooftop Equipment
Based on the above visible light image and infrared image analysis methods, a long lifecycle monitoring and management method for rooftop equipment is proposed. By establishing rooftop equipment archives, collecting basic equipment operating parameters, and combining UAV drone flight monitoring data, the equipment temperature change curve is plotted. Within the long lifecycle range, by analyzing the curve change trend and the change in equipment appearance, the equipment operation status is dynamically monitored, reducing the possibility of fire accidents caused by rooftop equipment anomalies.
Application Practice
Solution Overview
The above method was applied to an LCD factory in Anhui. The rooftop building area is about 260,000 m2, with 13.46 MW capacity photovoltaic panels installed. At the same time, exhaust gas combustion treatment devices, air cooling treatment devices, and other equipment are constructed on the rooftop.
The smart inspection platform based on UAV drones in this paper includes software and hardware modules. Among them, the UAV drone flight control system and gimbal camera together constitute a data acquisition hardware platform. Combined with automatic route planning technology, they form a data acquisition subsystem. The anomaly identification and localization module constitutes the data analysis subsystem.
The DJI Matrice 30T was used as the UAV drone flight platform. This flight platform is based on Real-Time Kinematic (RTK) technology and can achieve centimeter-level precise positioning. The single-flight endurance exceeds 41 minutes, providing effective guarantee for efficient and reliable task execution. The platform is equipped with the DJI Zenmuse H30T gimbal camera, which consists of a wide-angle camera, a zoom camera, an infrared camera, and a laser ranging sensor. Among them, the visible light zoom camera has an effective resolution of 20 million pixels and can achieve 50x optical hybrid zoom. The infrared imaging camera has a resolution of 640×512 pixels, with thermal sensitivity less than 50 mK, which can well meet the image quality requirements of the task. Using this gimbal camera, visible light and infrared images can be obtained simultaneously, providing convenience for anomaly localization and identification.
An adaptive cruise route was established, as shown in the figure. The designated inspection area is about 220,000 m2, with a planned total route length of more than 6,300 m. The expected total task duration is about 0.5 hours, and the data acquisition task can be completed within one flight.
According to the actual operating status of the equipment, a UAV drone flight visible light image safety checklist was established, as detailed in Table 1. According to the basic equipment operating parameters, an equipment operating temperature control checklist was formulated, as shown in Table 2.
| Item | Content |
|---|---|
| General | No abnormal hot spots in infrared image |
| Photovoltaic Panel | Average temperature of photovoltaic panel less than 90 °C |
| Temperature of photovoltaic combiner box less than 80 °C | |
| Exhaust Gas Treatment Device | Outer surface temperature of air duct less than 60 °C |
| Outer surface temperature of combustion furnace less than 80 °C | |
| Air Conditioning Device | Motor operating temperature less than 60 °C |
According to the local gray level analysis method, the actual infrared image calculation found that there was an anomaly point in the exhaust gas combustion treatment equipment on the rooftop of Building 2 of a certain LCD factory. By establishing rooftop equipment archives and taking the photovoltaic panels on the rooftop of Building 1 of a certain LCD factory as an example, the temperature change curve of the photovoltaic panels was plotted. It was found that the average temperature of the photovoltaic panels on the rooftop of Building 1 in the past month was 44.7 °C, which did not exceed the photovoltaic panel temperature control standard, but showed an upward trend. The operation and maintenance management personnel confirmed on-site that it was affected by the summer ambient temperature, causing the photovoltaic panel temperature to rise. Subsequent observation should be continued to ensure that the panel temperature remains below the standard value of 90 °C.
Benefit Analysis
Through the application of the above UAV drone inspection platform, a daily UAV drone inspection plan was formulated. The benefits generated are mainly in two aspects: one is the saved manpower inspection cost, and the other is the safety benefits from discovered problems and hazards. The benefits are calculated in detail as follows.
Manpower Benefit
$$T = S \times t \tag{3}$$
Where:
- T — total man-hours required for rooftop inspection of a certain LCD factory (h)
- S — rooftop inspection area of the certain LCD factory, taken as 220,000 m²
- t — time required for maintenance personnel to inspect 10,000 m² of rooftop (h)
For manual inspection, maintenance personnel need about 1 hour to inspect and measure temperature on 10,000 m² of rooftop. Substituting the above data into equation (3), the manual inspection time T_manual = 22 h. For UAV drone inspection, maintenance personnel need about 1 minute (0.0167 h) to inspect and measure temperature on 10,000 m² of rooftop. Substituting the data gives T_UAV = 0.35 h, which improves efficiency by 98.4% compared to manual inspection.
Safety Benefit
Through the use of the visible light image safety checklist (Table 1) and equipment operating temperature control checklist (Table 2) in the certain LCD factory, an average of 15 hazards were found per month. Typical hazards include: broken photovoltaic panel on the rooftop of Building 3, which was repaired and replaced to avoid heat accumulation leading to a fire; the outer surface temperature of the exhaust gas combustion treatment device on the rooftop of Building 2 was 131 °C, and after on-site confirmation by maintenance personnel, the device had a potential fire hazard, which was avoided by stopping the machine for repair and replacement; and some general industrial exhaust gas discharge air ducts on the rooftop of Building 1 had rust and damage at the top, which were replaced after on-site confirmation by maintenance personnel, effectively extending the overall service life of the air ducts.
Such safety benefits are not detectable due to the limitations of manual inspection perspective, and many possible fire accidents have been avoided. The indirect economic benefit depends on the value of the factory equipment and the production loss caused by fire, which is immeasurable and huge.
Future Prospects
The smart inspection platform based on UAV drones, in addition to the above functions and applications, has certain functional expandability in the later stage. For example, combined with AI technology, its intelligence level can be improved. The algorithm can intelligently identify abnormal points in the dual-spectrum images of UAV drone inspection, output a list of abnormal points and reports, further improving safety efficiency. For common peripheral construction operations inspections in factories, such as high-altitude maintenance operations and hot work, the high-resolution camera can intelligently monitor whether construction personnel have violations such as not wearing safety belts during high-altitude operations, and send corresponding alarms to safety management personnel. For possible chemical leakage accidents in factories, areas that emergency detection personnel cannot reach can be scanned and inspected. The inspection can also find cracks in the building body, improving the safety of the building itself.
Conclusion
- This paper proposes an application method for a smart factory inspection platform based on UAV drones, analyzes the feasibility and advantages of UAV drone smart inspection platforms in terms of efficiency and perspective, proposes anomaly point analysis methods under visible light and infrared images, and a long-lifecycle equipment status monitoring and management method. The method was applied in an LCD factory in Anhui.
- The application results show that the smart inspection platform based on UAV drones finds 15 hazards per month and improves inspection efficiency by 98.4%. It has high feasibility and economic benefits, and has expandability in functions such as AI recognition and construction inspection.