Pipe inspection system, pipe inspection method, pipe inspection program, and autonomous mobile robot for pipe inspection.

Abstract

It enables efficient and safe inspection and monitoring of piping, and allows for the acquisition of diverse data, including chemical and microbiological data, in real time and over a wide area, enabling early detection of abnormalities. [Solution] The present invention relates to a pipe inspection system S that utilizes a vehicle body 11 (autonomous mobile robot 1 for pipe inspection) capable of traveling inside pipes P. The vehicle body 11 is equipped with a drive motor 7 and a battery 6 to enable movement, and is also equipped with a LiDAR 21 that acquires point cloud data obtained by 3D scanning, a camera 22 that acquires image data obtained by imaging the inside of pipes, a biosensor 23 that detects microorganisms and chemical substances, a water quality sensor 24 that detects fluid characteristics, and a positioning sensor 25 that acquires position and attitude data of the vehicle body 11. A real-time analysis unit 31 analyzes this data and the detection results to determine whether or not there is an abnormality, and if there is an abnormality, an alarm is issued by an alarm issuance unit 32.

Description

Technical Field

[0001] The present invention relates to a pipeline inspection technology using a robot that autonomously travels inside pipelines such as sewer pipes, and to a pipeline inspection system, a pipeline inspection method, a pipeline inspection program, and an autonomous traveling robot for pipeline inspection, which are used to monitor the state of pipelines and the quality of fluids flowing inside the pipes.

Background Art

[0002] The maintenance and management of sewer pipelines is an important task carried out for the purpose of ensuring the proper operation of public infrastructure, maintaining the sanitary environment, and preventing floods. Conventionally, inspections and monitoring inside sewer pipes have been carried out by the following methods. (1) Regular manual inspection This is a method in which workers directly enter the sewer pipe and visually inspect the state inside the pipe. With this method, it is possible to confirm pipeline damage, deterioration, and the presence or absence of foreign objects. (2) Inspection by CCTV camera This is a method of acquiring images inside the pipe using a robot equipped with a small camera. The acquired images are later analyzed and used for evaluating the state of the pipeline (for sewer robots, see, for example, Non-Patent Document 1). (3) Fixed-point monitoring This is a method of installing sensors fixed at specific locations and monitoring data such as water quality, flow rate, and temperature. This method is suitable for continuously monitoring the situation at specific locations.

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

Summary of the Invention

[0004] However, although these methods are widely used in the maintenance and management of sewer pipelines, each method has its own unique constraints and limitations. Specifically, conventional methods have the following problems and challenges: (1) To improve work efficiency and ensure safety Manual inspections are time-consuming and labor-intensive, and involve risks as workers enter sewer pipes. Therefore, there is a need for technology that reduces the labor-intensive nature of manual inspections and ensures the safety of workers. (2) Acquisition of diverse data Inspections using CCTV cameras are limited to acquiring only video data and cannot obtain detailed information such as chemical or microbial data. Therefore, there is a need for technology that can acquire chemical and microbial data in addition to the video data from CCTV cameras. (3) Request for real-time and wide-area monitoring Fixed-point monitoring only provides data from specific locations, making it difficult to monitor the entire sewer pipeline network in real time over a wide area. Therefore, a system is needed that overcomes the limitations of fixed-point monitoring and enables real-time monitoring of the entire sewer pipeline network. (4) Early detection of abnormalities Even if an anomaly occurs, it cannot be detected early if it is outside the range of regular inspections or fixed sensors. Therefore, monitoring technology is needed to quickly identify the location of the anomaly and enable early response.

[0005] To address the above challenges, there is a need for the development of multi-functional monitoring technology that can efficiently and safely inspect and monitor sewer pipes, acquire real-time data over a wide area, and enable early detection of abnormalities. This invention has been made in view of the above circumstances, and its main objective is to provide a pipe inspection system, pipe inspection method, pipe inspection program, and autonomous mobile robot for pipe inspection that can efficiently and safely inspect and monitor sewer pipes, acquire diverse data including chemical and microbial data in real time and over a wide area, and enable early detection of abnormalities. [Means for solving the problem]

[0006] To achieve the above objectives, the pipe inspection system according to the present invention is It has a vehicle body that can travel inside the pipes, and this vehicle body has, A power mechanism for moving the vehicle body, The energy source that drives this power mechanism, A point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning the inside of a pipe, An image acquisition unit that acquires image data obtained by imaging the inside of the aforementioned pipe, A biosensor for detecting microorganisms or chemical substances in the aforementioned piping, A fluid characteristic sensor for detecting the fluid characteristics within the piping, The vehicle body is equipped with a positioning sensor that acquires position and orientation data of the vehicle body, The vehicle body or outside the vehicle body, A real-time analysis unit analyzes the acquired 3D point cloud data, the image data, and the position / orientation data, along with the detection results from the biosensor and the detection results from the fluid characteristics sensor to determine whether or not there is an abnormality. An alarm issuing unit that issues an alarm along with location information when the real-time analysis unit determines that an abnormality exists, It is characterized by possessing the following features.

[0007] Therefore, by 3D scanning the inside of the pipe using a point cloud data acquisition unit installed on the vehicle body that travels through the pipe, it is possible to obtain a detailed 3D model of the inner surface and internal structure of the pipe. This 3D model is used not only to accurately grasp the overall structure of the pipeline, but also to identify damaged or deteriorated areas and to analyze abnormal shapes inside the pipe. Furthermore, by combining it with high-resolution image data acquired by an image acquisition unit, it becomes possible to overlay visual information onto the 3D model and intuitively identify abnormal areas.

[0008] Furthermore, by incorporating biosensors that detect chemical substances and microorganisms within the pipes, the composition and contamination status of the fluids flowing through the pipes can be monitored in real time. This information is used not only for detecting structural abnormalities but also for monitoring water quality and predicting the spread of contamination. In addition, fluid characteristic sensors continuously monitor the physical properties of the fluids flowing through the pipes, such as pressure, flow rate, temperature, and viscosity, allowing for accurate understanding of fluid behavior within the pipeline and the identification of abnormal fluid movements and leak locations.

[0009] The positioning sensor plays a role in acquiring precise position and attitude data of the vehicle body, not only allowing for monitoring the robot's progress within the pipeline, but also providing specific location information when an anomaly is detected. This enables efficient repair work and maintenance planning for the affected area. Furthermore, the real-time analysis unit integrates the diverse data acquired from these sensors and instantly determines whether or not there is an anomaly. If an anomaly is detected, the alarm issuing unit issues an alarm including its location information, enabling a system to be established that prompts relevant parties to take swift action.

[0010] In order to accurately determine the presence or absence of anomalies in the real-time analysis unit, it is advisable to input the input data, including 3D point cloud data, image data, detection results from biosensors, and detection results from fluid characteristic sensors, into a machine learning model that has been pre-trained to determine the correlation between the presence or absence of anomalies and the input data, thereby determining the presence or absence of anomalies. Furthermore, in order to improve the accuracy of determining whether or not there is an abnormality in the piping, the vehicle body may be further equipped with a vibration sensor that detects whether or not there is an abnormality in the vibration pattern of the piping, and an acoustic sensor that detects whether or not there is damage to the piping. Furthermore, in order to ensure smooth movement of the vehicle body, the vehicle body may be further equipped with ultrasonic sensors that detect obstacles along its planned trajectory. Furthermore, the real-time analysis unit and alarm generation unit may be installed separately from the vehicle body, or installed within the vehicle body, depending on the purpose. Various variations can be set regarding whether either or both are installed within the vehicle body or installed independently of the vehicle body. [Effects of the Invention]

[0011] As described above, according to the pipe inspection system, pipe inspection method, pipe inspection program, and autonomous mobile robot for pipe inspection according to the present invention, Equipped with a power mechanism and energy supply within the vehicle body, it can autonomously travel inside pipes, enabling efficient and safe inspection and monitoring even in narrow sewer pipes where conventional manual labor or simple equipment would be difficult. This reduces the burden on workers, improves safety, and increases the efficiency of inspection work. Furthermore, it is equipped with various sensors such as a point cloud data acquisition unit, an image acquisition unit, a biosensor, a fluid characteristic sensor, and a positioning sensor. The 3D point cloud data, image data, position / orientation data, detection results of microorganisms and chemical substances, and fluid characteristic detection results acquired from these sensors can be analyzed in real time by the analysis unit. This allows for the acquisition of diverse data, including not only the physical structure inside the piping but also chemical and microbial data, in real time and over a wide area. This enables comprehensive monitoring of the piping condition, such as identifying damaged or deteriorated areas, understanding contamination levels, and detecting abnormal fluid behavior. Furthermore, if an abnormality is detected, the alarm unit can issue an alarm along with its location information, enabling early detection of the abnormality and allowing relevant parties to take swift countermeasures. This prevents the problem from escalating and improves the reliability of piping maintenance. Therefore, according to the present invention, efficient and safe inspection and monitoring inside sewer pipes are realized, various data including chemical data and microbial data can be acquired in real time and over a wide range, and early detection of abnormalities becomes possible.

Brief Description of the Drawings

[0012] [Figure 1] (a) is a block diagram showing the overall configuration of the pipe inspection system (fully external processing type) according to the present invention, and (b) is a perspective view showing the outline of the autonomous traveling robot for pipe inspection moving inside the pipe. [Figure 2] (a) is a flowchart showing the main processes for forming a learning model, and (b) is a diagram showing a configuration example of the machine learning device used in the present invention. [Figure 3] This is a flowchart for explaining the control flow of the pipe inspection system S according to the present invention. [Figure 4] This is a block diagram showing the overall configuration of the pipe inspection system (fully autonomous type) according to the present invention. [Figure 5] This is a block diagram showing the overall configuration of the pipe inspection system (partially external processing type) according to the present invention. [Figure 6] This is a block diagram showing the overall configuration of the pipe inspection system (hybrid type I) according to the present invention. [Figure 7] This is a block diagram showing the overall configuration of the pipe inspection system (hybrid type II) according to the present invention.

Embodiments for Carrying Out the Invention

[0013] Hereinafter, embodiments of a pipe inspection system using an autonomous traveling robot inside a pipe according to the present invention will be described based on the drawings.

[0014] In FIG. 1(a), the configuration of the pipe inspection system S according to the present invention is shown. In this example, the pipe inspection system S is shown inspecting a sewer pipe P. This pipe inspection system S transmits data collected by an autonomous mobile robot 1 moving inside the sewer pipe P to a real-time analysis unit 31 via a communication network 10. The real-time analysis unit 31 analyzes the data and, if necessary, issues an alarm via an alarm issuance unit 32.

[0015] The autonomous mobile robot 1 moves autonomously inside the sewer pipe P and monitors the condition of the pipe wall and the inside of the pipe in real time. It is equipped with a power unit 2, a driving control unit 3, a sensor unit 4, and a data transmission unit 5, which work together to enable advanced monitoring and data collection.

[0016] 1.Power part First, let's describe the power unit 2. The power unit 2 comprises the following elements. (1) Battery 6 Role: Responsible for supplying the overall energy to the vehicle body. Specifications: High-capacity battery (e.g., lithium-ion battery) enabling long-term continuous operation. (2) Drive motor Role: Controls the movement of the vehicle body. Specifications: High torque, low speed motor, suitable for use in sewer environments. (3) Caterpillar C / Wheels Role: A mechanism for movement within sewer pipes. Specifications: Non-slip material and design allow for smooth movement even in narrow spaces.

[0017] As shown in Figure 1(b), the autonomous mobile robot 1 for pipe inspection is equipped with a running mechanism at the bottom of the vehicle body 11. In this example, the running mechanism is composed of caterpillar tracks C, which include guide wheels 12 provided on both sides of the front of the vehicle body 11, drive wheels 13 provided on both sides of the rear of the vehicle body, a running belt 14 stretched between the front and rear opposing guide wheels 12 and drive wheels 13, and lower road wheels 15 pivotally attached to the base of the vehicle body 11 between the guide wheels 12 and drive wheels 13, and positioned to support the underside of the running belt 14 from the inside of the belt. While the example shown is a tracked vehicle (crawler-type vehicle) using a belt mechanism, it can be changed to a wheeled vehicle using multiple wheels as needed. These caterpillar tracks C and wheels should ideally be made of non-slip materials and have a non-slip design (shape) to allow for smooth movement even in the narrow spaces inside pipes.

[0018] The left and right caterpillar tracks C are powered by a drive motor 7 (power mechanism) located inside the vehicle body 11. The rotational power of the drive motor 7 is transmitted to the caterpillar tracks C via a sprocket (not shown), thereby driving them. The drive motor 7 is powered by electricity supplied from a battery 6 (energy source) and is located side by side inside the vehicle body 11, independently corresponding to the left and right caterpillar tracks C. This allows the vehicle body 11 to move in any direction (forward, backward, left, right, diagonally) within the piping. The battery in the power unit 2 also supplies power to the sensor unit 4, data transmission unit 5, and driving control unit 3, which will be described later. Furthermore, if the lights 16, real-time analysis unit 31, and alarm issuing unit 32 are located inside the vehicle body 11, this battery also supplies power to them. For this reason, a high-capacity battery such as a lithium-ion battery that enables long-term continuous operation is desirable. Additionally, the drive motor 7 should be a high-torque, low-speed motor to adapt to the environment inside the sewer.

[0019] 2. Driving Control Unit Next, the driving control unit 3 will be described. The driving control unit 3 is an integrated control system for realizing autonomous driving of the vehicle body, and mainly has three functions: navigation, location information and motion tracking, and map generation. (1) Navigation Role: Controls the position and direction of the vehicle body. Specifications: Autonomous driving algorithm, obstacle detection function (2) Location information and motion tracking (using GPS / IMU data) Role: Location and motion tracking Specifications: A system combining high-precision GPS and an inertial measurement unit (IMU). (3) Map generation Role: Generates maps of sewer pipes using 3D mapping data. Specifications: Real-time 3D mapping software, LiDAR is used.

[0020] The details of each function are explained below. (About navigation) The navigation system accurately determines the vehicle's current position and direction of travel, and calculates the optimal route to the destination. This system incorporates autonomous driving algorithms and obstacle detection capabilities, ensuring the vehicle can travel safely without external interference. It demonstrates particularly high reliability in confined spaces and environments with numerous obstacles. (Regarding location information and motion tracking) To accurately track the position and movement of the vehicle body, the positioning sensor (GPS / IMU) in the sensor unit 4, described later, is used. By combining a high-precision global positioning system (GPS) and an inertial measurement unit (IMU), position information is acquired in real time, and the acceleration and rotation angle of the vehicle body 11 are accurately measured. This enables stable position estimation even in environments where GPS signals are blocked. (Regarding map generation) Map generation is a function that creates a detailed 3D map of the inside of a pipe in real time, even in special environments such as inside a sewer pipe P. By using LiDAR (described later) and real-time 3D mapping software, three-dimensional data of the inside of the pipe and the pipe environment is acquired and processed with high precision. This function accurately projects the current position of the vehicle body 11 onto the map, enabling more efficient route selection and obstacle avoidance. By integrating these functions, this system supports the autonomous driving of the vehicle body 11 with high precision even in complex and dynamic environments, enabling safe and efficient work.

[0021] 3. Sensor section Next, we will explain the sensor unit 4. Sensor unit 4 is installed to acquire environmental information inside sewer pipes with high accuracy and from multiple perspectives, and mainly consists of three main components: an MMS unit (Mobile Mapping System unit), a biosensor, and a water quality sensor. These sensors complement each other to enable detailed data collection. (1) MMS unit (Mobile Mapping System) A mobile mapping system (MMS) will be used to perform 3D mapping of the inside of sewer pipes. Role: 3D mapping of sewer pipes Specifications: LiDAR, camera (2) Biosensors A sensor that acquires water quality and microbial data. Role: To collect water quality and microbial data. Specifications: Sensor for detecting specific chemicals and microorganisms (3) Water quality sensor Sensors that measure chemical substances, pollutants, pH levels, etc. Role: Measures chemical substances, pollutants, pH levels, etc. Specifications: Multifunctional water quality sensor (pH, dissolved oxygen, temperature, conductivity, etc.)

[0022] The details of each component are described below. (MMS Unit: Mobile Mapping System Unit) The MMS unit is the core system for realizing 3D mapping of sewer pipes. It incorporates a LiDAR 21 and a camera 22, and performs the following functions: (a) LiDAR: Performs a 360-degree omnidirectional laser scan to obtain 3D distance data by scanning the surrounding environment with high precision. This makes it possible to understand the physical structure inside the sewer pipe and details of damaged areas. (b) Camera This high-resolution camera is capable of capturing 360-degree panoramic images and acquiring video data from inside pipes. In particular, it can complement measurement data from LiDAR21 and be used as a visual record. In this example, the MMS unit (LiDAR 21 and camera 22) is integrated into the upper surface of the vehicle body 11. This arrangement allows for wide-area and detailed monitoring of the environment inside the pipes. The cooperation of these two sensors enables detailed monitoring of the sewer pipes and the generation of high-precision 3D maps. Note that the LiDAR 21 and camera 22 may be provided as separate sensors. For example, multiple cameras 22 may be arranged around the LiDAR 21. (About biosensors) The biosensor 23 plays a role in acquiring water quality and microbial data. Equipped with sensors to detect specific chemical substances and microorganisms, it enables the following environmental monitoring: Identify the types and concentrations of microorganisms in sewage and assess the level of contamination. We provide data to improve the efficiency of wastewater treatment by detecting chemical substances and hazardous materials. (Regarding water quality sensors) The water quality sensor 24 is a multi-functional sensor that measures the chemical and physical properties of wastewater. It acquires the following indicators in real time: (a) pH value: Measures the acidity or alkalinity of sewage. (b) Dissolved oxygen (DO): An indicator used to understand microbial activity and water quality. (c) Temperature: Measure the thermal environment in the sewage. (d) Conductivity: The concentration of ions in the water is measured to estimate the concentration of pollutants. These data serve as important indicators for water quality management and environmental assessment. The integration of these sensor units 4 makes it possible to accurately grasp not only the physical structure, damaged areas, and locations of obstacles within sewer pipes, but also environmental and water quality data, greatly contributing to the efficiency of maintenance work and environmental assessments.

[0023] (Regarding positioning sensors) Furthermore, the sensor unit 4 is also equipped with a positioning sensor (GPS, IMU) 25 that acquires position and attitude data of the vehicle body and collects positioning data to be used by the driving control unit 3. (a)GPS (Global Positioning System) GPS is a device that accurately determines the position of the vehicle body 11. It is installed on the top surface of the vehicle body 11 and is used to associate data collected by LiDAR 21 and camera 22 with its precise geospatial location. This ensures accurate location information of the collected data. In other words, it is used to map the collected data to a precise geospatial location. By measuring the latitude, longitude, and altitude of the vehicle body 11, geographic information and data are accurately linked. The data that can be collected by GPS primarily includes the following: (a) Location information (b) Precise geographical coordinates assigned to data collected by LiDAR or cameras.

[0024] (b)IMU(Inertial Measurement Unit) The IMU is a device that measures the movement of the vehicle body. It is installed inside the vehicle body and is used to correct the position of data acquired by the LiDAR 21 and camera 22. In other words, by measuring the acceleration and angular velocity of the mobile mapping vehicle, it plays a role in assisting the alignment of the data, thereby improving the accuracy of the collected data. The data that can be collected by the IMU mainly includes the following: (a) Information on the movement and posture of the vehicle body (b) Acceleration and rotation rate of the vehicle body: These are recorded and used for geographical correction of the data.

[0025] 4. Data transmission section Next, we will explain the data transmission unit 5. The data transmission unit 5 has the function of transmitting various data acquired by various sensors on the vehicle body 11 in real time to the real-time analysis unit 31 of the control center 30, and supports both wireless and wired communication. Role: Transmits acquired data to the control center 30 in real time. Specifications: High-bandwidth wireless communication systems (Wi-Fi, 5G), wired communication (fiber optic).

[0026] The details are explained below. The data transmission unit 5 is responsible for efficiently and reliably transmitting sensor data collected by various sensors mounted on the vehicle body 11 to the control center 30. This real-time communication allows the control center 30 to immediately grasp the status of the vehicle body 11, which is located remotely, and send instructions as needed. In addition, it enables immediate analysis and recording of acquired data, improving the operational efficiency of the vehicle body 11. The data transmission unit 5 is equipped with the following communication technologies. Wireless communication: High-bandwidth wireless communication technology is employed to achieve real-time data transmission between the vehicle body 11 and the control center. The following communication technologies are mainly used. Wi-Fi: A wireless communication technology suitable for medium-range communication, enabling high-speed and stable data transmission. 5G: A next-generation communication technology that enables ultra-high-speed data communication and low latency, and is particularly suitable for transmitting large amounts of data. This makes it possible to transmit high-resolution data and 3D mapping data in real time. Wired Communication: This system utilizes fiber optic technology for wired communication, offering both high-speed transmission of large amounts of data and high reliability. It is particularly effective in situations requiring stable communication over extended periods or in environments where high data security is essential. The data transmission unit 5 ensures the stability and efficiency of data transmission in any environment by switching between wireless and wired communication as needed. Furthermore, the transmitted data is analyzed and recorded in real time, allowing the control center 30 to make quick decisions and take action based on the information obtained. This system plays a crucial role in significantly improving the operational efficiency and safety of the vehicle body 11, as well as strengthening the monitoring capabilities of the control center 30, by enabling the immediate sharing and utilization of collected data.

[0027] 5. Real-time Analysis Unit Next, the real-time analysis unit 31 will be described. Role: Analyze acquired data in real time and detect anomalies. Specifications: AI / machine learning algorithms, database management systems. The real-time analysis unit 31 analyzes the data transmitted from the sensor unit 4 via the data transmission unit 5 in real time and detects the presence or absence of abnormalities. By utilizing AI and machine learning technologies, the real-time analysis unit 31 processes large amounts of diverse data quickly and accurately, thereby enabling early detection of abnormalities and improving the overall safety of the system.

[0028] The details are explained below. The real-time analysis unit 31 is responsible for immediately analyzing the data acquired by the sensor unit 4 and detecting abnormal patterns. This function allows the alarm issuance unit 32 to promptly issue an alarm when an abnormality is detected, prompting relevant parties to take appropriate action. This minimizes operational troubles and environmental risks.

[0029] The real-time analysis unit 31 uses a learning model to determine whether or not an anomaly is present. The main process for forming this learning model is as follows, as shown in Figure 2(a). (1) Data collection (Step S41) To form a learning model for determining the presence or absence of anomalies, the following data will be used as input data for the learning phase. (a) 3D point cloud data obtained by 3D scanning the inside of the pipe using LiDAR21: Data used to detect changes in the physical structure and shape of the sewer pipe (cracks, distortion, defects, etc.) (b) Image data obtained by imaging the inside of the pipe with camera 22: data for capturing visually recognizable abnormalities such as cracks, corrosion, and deterioration of the paint on the inner surface of the pipe. (c) Detection results from biosensors that identify the types and concentrations of microorganisms in sewage and detect the level of contamination and the presence or absence of chemical substances and hazardous materials. (d) Detection results from water quality sensors that measure the chemical and physical properties of wastewater (pH value, dissolved oxygen (DO), indicators for understanding microbial activity and water quality, temperature, conductivity, etc.), (2) Data preprocessing (Step S42) The acquired data may contain noise or missing values ​​in its raw form, so the following preprocessing is performed. (a) Noise filtering: Removes unwanted data fluctuations. (b) Data normalization: Unify the scale of the data to improve the accuracy of the analysis. (c) Imputation of missing values: Imputing missing data to ensure consistency of the analysis. (3) Feature extraction (Step S43) Key features are extracted from the preprocessed data. For example, these features may include: (a) Abnormal pH values ​​or temperature changes. (b) Rapid fluctuations in microbial or chemical concentrations. This improves the accuracy of anomaly detection. (4) Model learning (Step S44) To distinguish between normal and abnormal data, training is performed using a machine learning device 50. That is, data acquired from LiDAR 21, camera 22, biosensor 23, and water quality sensor 24 may be combined, and the presence or absence of abnormalities may be determined by a machine learning model. The machine learning device 50 is configured as a device connected to the communication network 10, for example, as shown in Figure 2(b), and includes an input data acquisition unit 51 that acquires data sets obtained from 3D point cloud data generated by 3D scanning the inside of a pipe (inner wall and scattered artifacts inside the pipe, etc.) with a LiDAR 21, image data generated by imaging the inside of the pipe with a camera 22, detection results detected by a biosensor 23, and detection results detected by a water quality sensor 24 as input data; a label acquisition unit 52 that acquires data sets including the presence or absence of abnormalities (binary classification of normal or abnormal) as labels; and a learning model construction unit 53 that constructs a learning model 55 by performing supervised learning using the input data and label pairs as training data. Here, the input data to be input to the input data acquisition unit 51 may be input directly from each of the sensors via the communication network 10, or the data from each sensor may be temporarily stored in a data storage unit (not shown) of the control center 30, and this stored data may be used as input data. Furthermore, the presence or absence of abnormalities in the sewer pipe may be determined using a learning model 55 constructed based on various data (D point cloud data, image data, and detection results from biosensors and water quality sensors) acquired by various sensors mounted on the mobile body M, and the determined result (presence or absence of abnormalities) may be displayed on a display unit (not shown) of the control center 30.

[0030] (Regarding the determination of whether or not there is an abnormality) The determination of whether or not there is an abnormality is performed specifically as follows: First, we perform data preprocessing and feature extraction. LiDAR data: Point cloud data is converted into a 3D mesh, and geometric features indicating cracks or surface anomalies (e.g., abrupt changes in distance, surface irregularities) are extracted. Camera data: Features such as edges, cracks, and corrosion patterns are extracted from images using convolutional neural networks (CNNs), etc. Furthermore, if a multispectral camera is installed, abnormal reflection patterns or temperature anomalies at specific wavelengths may be detected at the pixel level and used as features. Next, the features obtained from each sensor are combined to generate a feature vector for finally classifying whether or not there is an anomaly.

[0031] The labels used to determine the presence or absence of abnormalities can be binary labels, such as "1" for abnormalities and "0" for normal conditions. In other words, a classification model such as a Support Vector Machine (SVM), Random Forest, or Deep Learning model (e.g., a multilayer perceptron or end-to-end CNN+RNN model) is used to perform a two-class classification of normal and abnormal. Then, a dataset containing both normal and abnormal conditions is prepared, and the learning model is trained.

[0032] In the above, the presence or absence of abnormalities is determined by a two-class classification (normal or abnormal), but labels indicating the specific nature of the abnormality (cracks, corrosion, deformation, abnormal pH value, abnormal temperature change, abnormal microbial concentration or chemical concentration) may also be used. This improves the accuracy of anomaly detection. Furthermore, in determining whether or not an anomaly exists, threshold-based detection may be used: Anomalies may be detected using a pre-set threshold. For example, anomalies may be identified by comparing changes in the shape of LiDAR data, color changes in RGB images, or temperature anomalies in thermal images with the threshold. Alternatively, a rule-based algorithm may be used to identify anomaly locations based on pre-defined rules. For example, an anomaly may be determined when a specific shape change or increase in density is detected.

[0033] 6. Alarm issuing unit Next, the alarm issuing unit 32 will be described. The alarm issuing unit 32 automatically issues an alarm when an abnormality is detected by the real-time analysis unit 31, and promptly notifies the relevant parties. This alarm issuing unit 32 plays a crucial role in accelerating problem resolution by consistently handling everything from early detection of abnormalities to information dissemination to the relevant parties. Role: Issues an alarm and notifies relevant parties when an anomaly is detected. Specifications: Mobile notification app, siren, light, email notifications.

[0034] The details are explained below. The main role of the alarm issuing unit 32 is to automatically trigger an alarm in response to detected anomalies and immediately notify the relevant parties. It also provides appropriate response instructions depending on the type and urgency of the anomaly. This enables the relevant parties to respond quickly and appropriately. (1) Functions and Specifications When the real-time analysis unit 31 detects an anomaly, the alarm issuing unit 32 automatically responds to this. An alarm will be triggered. The automatic alarm function will enable relevant personnel to respond to the anomaly in real time. (2) Alarm notification The alarm issuing unit 32 is equipped with various notification methods, enabling it to deliver information to relevant parties quickly and reliably. The main notification methods are as follows: Mobile notification app: Provides immediate notifications of anomalies and detailed information via smartphone. Email notification: Detailed anomaly reports and recommended countermeasures will be sent via email. Sirens and lights: Provide immediate attention by issuing visual and audible warnings at the scene. (3) Instructions for action The notification includes the type of anomaly, its urgency, and recommended actions. This allows stakeholders to quickly take the most appropriate response to the situation. For example, in the case of a minor anomaly, instructions would be given to continue monitoring, while in the case of a serious anomaly, an emergency shutdown or notification to a specialist team would be recommended. (4) Features The alarm unit comprehensively covers everything from anomaly notification to response instructions, and has the following features: Speed: Designed to minimize delays from anomaly detection to notification. Diversity: Provide multiple notification methods and deliver information in the most appropriate way based on the circumstances of each stakeholder. Reliability: Automated alarm system reduces the risk of missing anomalies. The alarm issuing unit 32 is a crucial element that supports the safety of the entire system, and by streamlining the process from anomaly detection to response, it enables rapid problem solving and improved operational efficiency.

[0035] 7. Control Flow Next, we will explain the control flow of this piping inspection system S. The control flow of this pipe inspection system S integrates a series of processes, starting from the activation of the power unit 2, to data acquisition, analysis, and alarms. This system improves operational efficiency and safety by monitoring the conditions inside sewer pipes in real time and quickly detecting and notifying of abnormalities. The control operation of the control flow consists of the following steps, as shown in Figure 3. (1) The power unit 2 drives the vehicle body, moving it through the sewer pipe (step S51). (2) The GPS / IMU acquires location information, and the navigation system controls the route (step S52). (3) The MMS unit performs 3D mapping to visualize the condition inside the sewer pipe (step S53). (4) Biosensors and water quality sensors monitor the quality of wastewater in real time (Step S54). (5) The data acquired by the data transmission unit 5 is transmitted to the control center 30 (step S55). (6) The real-time analysis unit 31 analyzes the data and detects anomalies (step S56). (7) If the alarm issuing unit 32 detects an abnormality, it issues an alarm and notifies the relevant parties (step S57).

[0036] The following describes each step of the control flow in detail. (Step S51; Power unit 2 moves the vehicle body and moves it through the sewer pipe) The system begins with the activation of the power unit 2. A battery mounted on the vehicle body supplies power to the motor, which drives the wheels or tracks. This mechanism allows the vehicle body to move stably through the sewer pipes. Furthermore, the driving control unit 3 manages the vehicle body's route based on navigation data, ensuring safe and efficient movement. (Step S52: GPS / IMU acquires location information, and the navigation system controls the route to the destination.) The vehicle's position information is acquired in real time by GPS and IMU (Inertial Measurement Unit). Based on this information, the navigation system calculates the route to the destination and controls the vehicle's direction of travel. GPS and IMU data are updated as needed, allowing the system to respond to dynamic environmental changes such as obstacles and sudden changes in direction. As a result, the vehicle can move smoothly even in complex terrain and narrow spaces within pipes. (Step S53: The MMS unit performs 3D mapping to visualize the condition inside the sewer pipe.) The MMS (Mobile Mapping System) unit plays a central role in visualizing the detailed conditions inside sewer pipes. The onboard LiDAR scans the surrounding environment and generates high-precision 3D maps in real time. Furthermore, high-resolution cameras acquire video data, complementing the LiDAR data to provide a more detailed understanding of the pipe's condition. This data is used to detect damage and abnormalities within the pipes at an early stage. (Step S54; Biosensors and water quality sensors monitor wastewater quality in real time) The biosensor and water quality sensor mounted in sensor unit 4 continuously monitor the quality of wastewater. The biosensor detects microorganisms and specific chemical substances to assess the level of contamination in the wastewater. Meanwhile, the water quality sensor measures indicators such as pH, dissolved oxygen, temperature, and conductivity to understand the water quality status in real time. This function allows for accurate monitoring of wastewater conditions, contributing to environmental protection and the realization of appropriate wastewater treatment. (Step S55: The acquired data is transmitted from the data transmission unit 5 to the control center.) The collected data is transmitted to the control center via the data transmission unit 5. Wireless communication (Wi-Fi or 5G) is used for real-time data transmission, ensuring high-speed and stable communication. Wired communication (optical fiber) may also be used as needed, enabling reliable transmission of large amounts of data. (Step S56: The real-time analysis unit 31 analyzes the data and detects anomalies.) Data sent to the control center is immediately analyzed by the real-time analysis unit 31. This real-time analysis unit 31 uses AI and machine learning algorithms to detect anomaly patterns from the collected data. After preprocessing such as noise filtering and normalization, the data is feature-extracted and analyzed based on the model. This enables early detection of anomalies, allowing appropriate action to be taken before problems occur. (Step S57: If the alarm unit detects an abnormality, it issues an alarm and notifies the relevant parties.) If an anomaly is detected, the alarm unit 32 automatically activates. Relevant parties are immediately notified via a mobile notification app, email, siren, lights, etc. The notification includes the type of anomaly, its urgency, and recommended countermeasures, enabling a quick and appropriate response. This process plays a crucial role in ensuring operational safety and efficiency.

[0037] 8. Examples of applications for pipe inspection systems This system monitors the condition of sewer pipes in real time, detects anomalies, and enables rapid response. Its design ensures the preservation of sewer infrastructure, environmental safety, and efficient maintenance. To illustrate how this system is used, two scenarios—sewer pipe damage detection and water quality anomaly detection—are described below. These scenarios detail the system's flow and how each component works in conjunction.

[0038] (1) Scenario 1: Detection of damage inside sewer pipes The Intelligent Sewer Guardian is activated to detect damaged areas within sewer pipes. Power unit 2 activates, the vehicle enters the sewer pipe, and begins to move. While moving, the driving control unit 3 uses GPS and IMU to determine the vehicle's current position, and the navigation system controls the route to the destination. It proceeds safely and efficiently while detecting obstacles and conditions within the narrow pipe. While the vehicle body is moving, the sensor unit 4 operates, and 3D mapping is performed by the MMS unit. LiDAR scans the surrounding environment and generates a detailed 3D model of the inside of the sewer pipe. At the same time, biosensors and water quality sensors acquire environmental data inside the pipe in real time and monitor for any abnormalities. All of this data is aggregated within the vehicle body and transmitted in real time to the control center 30 via the data transmission unit 5. In the control center 30, the real-time analysis unit 31 analyzes the acquired data. Utilizing AI and machine learning algorithms, it scrutinizes the structural data inside the pipe to identify damaged areas. For example, it detects cracks and defects in the pipe wall with high precision and understands the extent of the damage and its impact in detail. If an abnormality is identified, the analysis results are immediately transmitted to the alarm issuance unit 32. When the alarm unit 32 detects damage, an alarm is immediately triggered. This alarm is communicated to relevant parties via a mobile notification app, email, and even on-site sirens and lights. The notification includes details of the damage, urgency, and recommended actions, enabling relevant parties to take swift and appropriate action.

[0039] (2) Scenario 2: Detection of water quality abnormalities The Intelligent Sewer Guardian is operated to detect water quality abnormalities within the sewer system. Power unit 2 is activated, and the vehicle enters the sewer pipe and begins to move. The driving control unit 3 uses GPS and IMU to determine the vehicle's position in real time, and the navigation system controls the route to the destination. Avoiding complex terrain and obstacles within the pipe, the vehicle proceeds towards its destination. While the vehicle is in motion, the sensor system operates, and biosensors and water quality sensors acquire water quality data. This allows for the measurement of concentrations of microorganisms and chemical substances in the wastewater, as well as pH levels, dissolved oxygen, temperature, and conductivity. For example, if an unusual pattern is detected, such as an abnormal concentration of chemical substances or a rapid change in pH, all of this data is collected within the vehicle body and transmitted in real time to the control center by the data transmission unit 5. Data arriving at the control center is analyzed by the real-time analysis unit 31. Using AI and machine learning algorithms, anomalies are detected by comparing the data with normal water quality data. For example, if the concentration of a specific harmful chemical substance exceeds a threshold, the source of the substance and the extent of its impact are immediately identified. When an anomaly is detected, the alarm unit 32 automatically activates and triggers an alarm. Relevant parties are notified via a mobile notification app, email, and on-site sirens and lights. The notification includes detailed information about the anomaly, its location, urgency, and recommended actions. This information allows relevant parties to take swift action and prevent the problem from escalating.

[0040] These scenarios demonstrate how the intelligent sewer guardian can closely monitor conditions within sewer pipes, enabling rapid identification and response to anomalies. The implementation of such a system is expected to significantly improve the maintenance and efficient operation of sewer infrastructure.

[0041] 9. Advantages of the pipe inspection system The pipe inspection system is an advanced system that dramatically improves the efficiency and accuracy of sewer management. By utilizing autonomous driving technology, a wide variety of sensors, and real-time analysis capabilities, this system offers many advantages that were not possible with conventional methods. These advantages are described in detail below. (1) Efficient operation The autonomous driving function of this system significantly reduces manual inspection work. Traditional sewer inspections required workers to directly enter the pipes, but the introduction of the Intelligent Sewer Guardian reduces the burden on workers and minimizes contact with hazardous environments. Furthermore, its ability to operate continuously for extended periods allows for efficient inspection of large areas of sewer pipes. This improvement in operational efficiency contributes to shorter working times and improved inspection accuracy. (2) Comprehensive data acquisition The pipe inspection system S enables 3D mapping using an MMS unit and simultaneous acquisition of environmental data using biosensors and water quality sensors. This allows for detailed and comprehensive information to be obtained, including not only the physical structure and damaged areas inside sewer pipes, but also water quality and microbial conditions. Furthermore, by integrating multiple sensors, comprehensive data collection regarding the sewer environment is achieved. This multifaceted data acquisition dramatically improves the accuracy of sewer management. (3) Improved real-time performance This system enables real-time data acquisition and analysis. Data acquired by sensors is analyzed immediately, and if an anomaly is detected, countermeasures can be taken quickly. This real-time capability allows for early detection of anomalies, enabling appropriate measures to be taken before problems escalate. Rapid response is a crucial element in maintaining the stability of sewage infrastructure. (4) Improved safety The intelligent sewer guardian has the ability to gain a detailed understanding of hazardous areas and environmental conditions within sewer pipes. This allows workers to develop safe work plans before directly entering dangerous areas. Furthermore, its autonomous navigation and remote control capabilities minimize the need for workers to work in high-risk environments. These features are crucial for ensuring worker safety and preventing accidents. (5) Cost reduction Early detection of abnormalities allows for appropriate action before major repairs become necessary. This reduces large-scale repair work and the associated long-term costs. Furthermore, efficient operation reduces the frequency of inspections and repairs, thus lowering the costs of regular maintenance. These cost reductions contribute to the long-term sustainability of sewer management. The Intelligent Sewer Guardian offers numerous advantages, including efficiency, comprehensive data acquisition, real-time analysis, improved safety, and cost reduction, setting a new standard for sewer management. The introduction of such technology is expected to significantly enhance the reliability and sustainability of sewer infrastructure.

[0042] 10. Arrangement of the real-time analysis unit and alarm issuance unit In the above, we have shown a configuration (fully external processing type) in which various data collected by the sensor unit 4 are transmitted to a real-time analysis unit 31, which is provided separately from the vehicle body, that is, installed independently of the vehicle body, for analysis, and based on the analysis results, an alarm is issued by the alarm issuance unit 32 as needed. (1) Fully external processing type This fully external processing type is configured such that the robot is only responsible for data collection, while the real-time analysis unit 31 and alarm issuance unit 32 are located externally (e.g., in a control center). In this system, the data acquired by the robot is transmitted externally in real time, and analysis and alarm issuance are performed externally. The advantage of this method lies in its ability to reduce the weight of the robot. This makes it possible to reduce the manufacturing cost and power consumption of the robot. In addition, advanced analytical capabilities can be utilized at an external control center. On the other hand, a communication infrastructure is essential, and it may be affected by communication delays. This configuration is suitable for urban areas with well-developed communication environments or situations where existing communication infrastructure can be used. The following variations are possible for the installation locations of the real-time analysis unit 31 and the alarm issuance unit 32. (2) Fully autonomous robots As shown in Figure 4, the real-time analysis unit 31 and alarm issuance unit 32 may be built into the vehicle body (fully autonomous). In this configuration, analysis and alarms are completed within the robot, so there is almost no dependence on external infrastructure. Therefore, it can operate even in environments where communication is difficult, such as deep underground or remote locations. This fully autonomous system has the advantage of not requiring external infrastructure and being unaffected by communication delays. On the other hand, it also has the disadvantages of a more complex robot design and increased power consumption required for system operation. This configuration is particularly effective in remote locations where communication is difficult or in situations where full autonomy is required. (3) Partial external processing type The partially external processing type, as shown in Figure 5, is configured such that the robot body is equipped with a real-time analysis unit 31, and the alarm issuance unit 32 is installed externally (for example, at a control center). In this system, the robot analyzes the anomaly and transmits the results to the external control center. Alarm issuance is performed by an external system. The advantage of this method is that it reduces the processing load on the robot itself. Also, since alarm management is centralized externally, monitoring and operation become more efficient. However, a disadvantage is that a communication environment is essential, and the system will cease to function if communication is interrupted. This configuration is suitable for situations where a stable communication environment is available and remote monitoring of analysis results is desired. (4) Hybrid Type I In the Hybrid Type I, as shown in Figure 6, the robot is equipped with a real-time analysis unit 31, and the alarm issuing unit 32 is distributed both inside and outside the robot. In this configuration, the robot analyzes abnormalities, and for minor alarms, the robot itself issues the alarm. On the other hand, for serious alarms, the control center can handle the situation. The advantage of this approach lies in its high degree of flexibility. Minor issues can be addressed immediately by robots, while critical issues are managed by a control center, ensuring rapid response. However, the system design is complex, requiring advanced knowledge for design and operation. This configuration is particularly effective in applications where safety is critical, such as detecting major damage in sewer systems. (5) Hybrid Type II In the Hybrid Type II, as shown in Figure 7, the real-time analysis unit 31 is installed independently of the vehicle body, and the alarm dispatch unit 32 is distributed both inside and outside the robot. This configuration combines the advantages and disadvantages of a fully external processing system and enables rapid identification of abnormal locations on-site.

[0043] 11. Adding other sensors In addition to the above-mentioned sensors, a vibration sensor 26 and an acoustic sensor 27 may also be provided. (1) About vibration sensors The vibration sensor 26 is a device that detects vibrations and motions and converts them into electrical signals. It is used for anomaly detection and condition monitoring in pipes, structures, and mechanical equipment. The vibration sensor detects vibrations as acceleration, velocity, or displacement, and converts this into analyzable data to identify abnormal locations and monitor deterioration. Vibration sensors are particularly effective in identifying damaged sections of sewer pipes. They can detect abnormal vibration patterns and identify potential damage or cracks. Furthermore, the results from vibration sensors can be integrated with information from other sensors (e.g., MMS unit and camera data) to enable more accurate analysis. (2) About acoustic sensors The acoustic sensor 27 is a device that uses sound waves to detect the internal condition of pipes and structures. By detecting changes in the propagation speed and reflection characteristics of sound waves, it identifies internal cracks, cavities, and leaks. It is widely used to detect damage and deterioration in sewer pipes. Acoustic sensors are characterized by their ability to analyze internal conditions non-destructively. Even in areas difficult to inspect visually, abnormalities can be identified by analyzing changes in sound waves. This characteristic is extremely useful in confined spaces and dark environments such as sewer pipes. The detection results from acoustic sensors can be combined with vibration sensors and visual data to achieve more reliable anomaly detection. Both the vibration sensor 26 and the acoustic sensor 27 are technologies that play an important role in sewer management. By appropriately utilizing the characteristics of each, it is possible to quickly and accurately identify abnormal areas and maintain the integrity of pipes and structures. (3) An ultrasonic sensor 28 may be provided in order to detect obstacles in the planned path of the robot moving through the sewer pipe. This ultrasonic sensor 28 contributes to the navigation of the robot, enabling it to travel efficiently and safely through the pipe and avoid obstacles. It is also used to measure fluid flow rate and water level, playing a role in monitoring the conditions inside the pipe in real time.

[0044] 12. Other functions of the autonomous mobile robot for pipe inspection (1) Backup mode A backup mode may be provided that automatically activates when a robot traveling through a sewer pipe becomes unable to move forward. This backup mode allows the robot to safely return to its original position even if it becomes stuck due to an obstacle, a corner of the pipe, or a joint. Specifically, when the robot detects an obstacle or an impediment to its path, it immediately stops moving and autonomously switches to backup mode. When backup mode is activated, the robot begins to move backward, returning to its original traversable position. This function allows the robot to avoid obstacles or recalculate its path and select a new route. Backup mode is designed to ensure the robot's safe and efficient operation even in the narrow spaces and complex terrain of sewer pipes. Furthermore, the backup mode not only operates based on the robot's autonomous capabilities but can also receive remote support from a control center as needed. This allows for flexible responses in situations where progress becomes impossible.

[0045] (2)Self-repair function A robot traveling through sewer pipes may be equipped with a self-repair function to temporarily repair any damaged or faulty areas it detects. This function activates when the robot detects minor damage or deterioration within the sewer pipe being inspected, and quickly performs repair work on the spot, thereby preventing major damage or further deterioration. Self-healing robots have built-in special repair agents and tools that they use to perform emergency repairs on detected damage. For example, they may fill small cracks with repair agents or temporarily seal leaks. This repair work maintains the functionality of the pipeline and prevents further damage until permanent repairs can be carried out. While the self-repair function can operate completely autonomously, more precise action can be achieved by combining it with remote control from a control center and monitoring of the repair process. This function is particularly useful in remote locations and environments that are difficult to access, improving the operational efficiency of sewer systems.

[0046] While the above description focused on inspection systems and self-propelled robots for inspecting sewer pipes, these methods are applicable not only to the location of sewer pipes but also to inspecting the inside of other types of piping. In such cases, the design specifications of the piping inspection system or robot need to be adjusted according to the target piping environment. [Explanation of Symbols]

[0047] 1. Autonomous mobile robot 7. Drive motor (power mechanism) 6. Battery (energy source) 11. Body 21. LiDAR (Point Cloud Data Acquisition Unit) 22 Camera (Image acquisition unit) 23 Biosensors 24. Water quality sensors (fluid characteristic sensors) 25 Positioning Sensor 26 Vibration Sensor 27 Acoustic Sensor 28 Ultrasonic Sensors 31 Real-time Analysis Department 32 Alarm issuing unit 55 Learning Models S Pipe Inspection System P piping

Claims

1. It has a vehicle body that can travel inside the pipes, and this vehicle body has, A power mechanism for moving the vehicle body, The energy source that drives this power mechanism, A point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning the inside of a pipe, An image acquisition unit that acquires image data obtained by imaging the inside of the aforementioned pipe, A biosensor for detecting microorganisms or chemical substances in the aforementioned piping, A fluid characteristic sensor for detecting the fluid characteristics within the piping, The vehicle body is equipped with a positioning sensor that acquires position and orientation data, The vehicle body or independently thereof, A real-time analysis unit analyzes the acquired 3D point cloud data, the image data, and the position / orientation data, along with the detection results from the biosensor and the detection results from the fluid characteristics sensor to determine whether or not there is an abnormality. An alarm issuing unit that issues an alarm along with location information when the real-time analysis unit determines that an abnormality exists, A piping inspection system characterized by comprising the following:

2. The determination of whether or not there is an abnormality in the real-time analysis unit is as follows: The pipe inspection system according to claim 1, characterized in that the system is performed by an abnormality determination unit that determines the presence or absence of an abnormality by inputting the input data, including the 3D point cloud data, the image data, the detection results from the biosensor, and the detection results from the fluid characteristic sensor, into a learning model that has been pre-trained to determine the correlation between the presence or absence of an abnormality and the input data.

3. The pipe inspection system according to claim 1, further comprising: a vibration sensor for detecting whether or not there is an abnormality in the vibration pattern of the pipe; and an acoustic sensor for detecting whether or not there is damage to the pipe.

4. The pipe inspection system according to claim 1, further comprising an ultrasonic sensor in the vehicle body for detecting obstacles along its planned trajectory.

5. The piping inspection system according to claim 1, wherein the real-time analysis unit is provided in a manner that does not belong to the vehicle body.

6. The piping inspection system according to claim 1, wherein the real-time analysis unit is provided inside the vehicle body.

7. It has a vehicle body that can travel inside the pipes, and this vehicle body has, A power mechanism for moving the vehicle body, The energy source that drives this power mechanism, A point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning the inside of a pipe, An image acquisition unit that acquires image data obtained by imaging the inside of the aforementioned pipe, A biosensor for detecting microorganisms or chemical substances in the aforementioned piping, A fluid characteristic sensor for detecting the fluid characteristics within the piping, A pipe inspection method comprising a positioning sensor that acquires position and attitude data of the vehicle body, and for inspecting the inside of the pipe, A real-time analysis step to determine whether or not there is an abnormality by analyzing the acquired 3D point cloud data, the image data, and the position / orientation data, as well as the detection results from the biosensor and the detection results from the fluid characteristics sensor. If an abnormality is determined in the aforementioned real-time analysis step, an alarm is issued along with the location information. A pipe inspection method characterized by comprising the following:

8. The determination of whether or not there is an abnormality in the real-time analysis step is as follows: The pipe inspection method according to claim 7, characterized in that it is performed by an abnormality determination step in which the presence or absence of an abnormality is determined by inputting the input data, including the 3D point cloud data, the image data, the detection result from the biosensor, and the detection result from the fluid characteristic sensor, into a learning model that has been pre-trained to determine the correlation between the presence or absence of an abnormality and the input data.

9. A pipe inspection program for causing a computer to perform each step of the pipe inspection method according to any one of claims 7 to 8.

10. It has a vehicle body that can travel inside the pipes, and this vehicle body has, A power mechanism for moving the vehicle body, The energy source that drives this power mechanism, A point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning the inside of a pipe, An image acquisition unit that acquires image data obtained by imaging the inside of the aforementioned pipe, A biosensor for detecting microorganisms or chemical substances in the aforementioned piping, A fluid characteristic sensor for detecting the fluid characteristics within the piping, A positioning sensor that acquires position and orientation data of the vehicle body, A data transmission unit transmits the acquired 3D point cloud data, image data, position and orientation data, detection results from the biosensor, and detection results from the fluid characteristic sensor to a real-time analysis unit installed independently of the vehicle body in order to analyze the acquired data and detection results and determine whether or not there is an abnormality. An autonomous mobile robot for pipe inspection, characterized by being equipped with the following:

11. In place of the aforementioned data transmission unit, A real-time analysis unit analyzes the acquired 3D point cloud data, the image data, the position and orientation data, the detection results from the biosensor, and the detection results from the fluid characteristics sensor to determine whether or not there is an abnormality. An alarm issuing unit that issues an alarm when the real-time analysis unit determines that an abnormality exists, The autonomous mobile robot for pipe inspection according to claim 10, characterized by comprising the above.

Images