Sanitation vehicle intelligent monitoring system

Abstract

The application provides a sanitation vehicle intelligent monitoring system, which comprises a front-end sensing module, a data transmission module and a back-end processing module.The front-end sensing module is used for collecting real-time data of the sanitation vehicle in real time, and the real-time data comprises running state data, driver behavior data and working environment data.The data transmission module is in communication connection with the front-end sensing module and is used for uploading the real-time data to the back-end processing module.The back-end processing module is in communication connection with the data transmission module and is used for processing and analyzing the received real-time data to identify abnormal states and generate control instructions.The application execution module is in communication connection with the back-end processing module and is used for executing early warning prompts or vehicle control operations based on the control instructions.The abnormal states at least include blind area obstacles and driver fatigue driving behaviors of the sanitation vehicle.In this way, the real-time identification and hierarchical intervention of abnormal states are realized by fusing driver behavior monitoring and blind area early warning, and the safety of sanitation operation and the response efficiency of management are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring technology, and in particular to an intelligent monitoring system for sanitation vehicles. Background Technology

[0002] Vehicle-mounted monitoring systems are widely used in vehicle management, using devices such as GPS and cameras to collect vehicle location and video footage. However, existing systems are mostly limited to simple driving recording and positioning, lacking in-depth perception of driver status and proactive early warning capabilities for operational scenarios. Especially in sanitation vehicle operations, vehicles often face safety hazards such as large blind spots and driver fatigue. Existing technologies cannot identify and intervene in abnormal behaviors in real time and at different levels, resulting in delayed accident response and failing to meet the urgent needs of smart sanitation for safety and intelligent management. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide an intelligent monitoring system for sanitation vehicles, which integrates driver behavior monitoring and blind spot warning to achieve real-time identification and graded intervention of abnormal states, thereby significantly improving the safety of sanitation operations and management response efficiency.

[0004] In a first aspect, embodiments of the present invention provide an intelligent monitoring system for sanitation vehicles. The system includes: a front-end sensing module, a data transmission module, a back-end processing module, and an application execution module. The front-end sensing module is used to collect real-time data from the sanitation vehicle. The real-time data includes: operating status data, driver behavior data, and operating environment data. The data transmission module is communicatively connected to the front-end sensing module and is used to upload the real-time data to the back-end processing module. The back-end processing module is communicatively connected to the data transmission module and is used to process and analyze the received real-time data to identify abnormal states and generate control commands. The application execution module is communicatively connected to the back-end processing module and is used to execute early warning prompts or vehicle control operations based on the control commands. Abnormal states include at least obstacles in the sanitation vehicle's blind spots and driver fatigue.

[0005] In a preferred embodiment of the present invention, the aforementioned front-end perception module includes: a vehicle status acquisition submodule, a driver status acquisition submodule, and an environment perception submodule; the vehicle status acquisition submodule is used to acquire the driving parameters and operating parameters of the sanitation vehicle; the driver status acquisition submodule is used to acquire the driver's facial features and behavioral posture; and the environment perception submodule is used to acquire obstacle information around the sanitation vehicle.

[0006] In a preferred embodiment of the present invention, the vehicle status acquisition submodule includes: a GPS locator, a speed sensor, and a cleaning efficiency detector; the GPS locator, speed sensor, and cleaning efficiency detector are all communicatively connected to the data transmission module; the driver status acquisition submodule includes: a DSM camera; the DSM camera is used to capture the driver's facial image in real time; the environmental perception submodule includes at least one blind spot monitoring radar or surround view camera; the blind spot monitoring radar or surround view camera is installed on the side or rear of the sanitation vehicle to detect visual blind spots around the vehicle body and obtain detection data.

[0007] In a preferred embodiment of the present invention, the aforementioned backend processing module includes: a data analysis submodule and an alarm decision submodule; the data analysis submodule is used to perform real-time analysis on the received real-time data based on a preset algorithm to determine whether there is an anomaly and obtain the analysis result; the alarm decision submodule is connected to the data analysis submodule and is used to generate a graded alarm command based on the analysis result.

[0008] In a preferred embodiment of the present invention, the above-mentioned data analysis submodule includes: a fatigue driving analysis unit, a blind spot warning analysis unit, and a data fusion unit; the fatigue driving analysis unit is used to receive image data from the DSM camera and detect the duration of the driver's eye closure based on the image data using an image recognition algorithm; the blind spot warning analysis unit is used to receive detection data from the environmental perception submodule and determine whether there are obstacles within a preset range of the sanitation vehicle; the data fusion unit is used to perform correlation analysis between the operating status data and the driver's behavior data to identify high-risk driving scenarios.

[0009] In a preferred embodiment of the present invention, the alarm decision submodule is configured with a first alarm threshold and a second alarm threshold, wherein: when the data analysis submodule determines that the abnormal state reaches the first alarm threshold, the alarm decision submodule generates a first-level alarm instruction, which is used to trigger the audible and visual prompts of the vehicle terminal; when the data analysis submodule determines that the abnormal state continues and reaches the second alarm threshold, the alarm decision submodule generates a second-level alarm instruction, which is used to trigger remote voice intervention or forced vehicle control.

[0010] In a preferred embodiment of the present invention, the data transmission module includes: a wireless communication unit and a data security unit; the wireless communication unit uses 4G / 5G or WiFi communication technology to establish a real-time data link between the front-end sensing module and the back-end processing module; the data security unit is connected to the wireless communication unit and is used to encrypt and compress the transmitted data.

[0011] In a preferred embodiment of the present invention, the application execution module includes: an on-board alarm unit and a vehicle control unit; the on-board alarm unit is installed in the cab of the sanitation vehicle and is used to issue visual or audible warnings based on the control of the back-end processing module; the vehicle control unit is communicatively connected to the electronic control unit of the sanitation vehicle and is used to perform operations such as deceleration, activating hazard lights, or forced stopping based on the control commands of the back-end processing module.

[0012] In a preferred embodiment of the present invention, the backend processing module is further configured to continuously receive feedback data from the frontend sensing module through the data transmission module after issuing the warning command. If no abnormal state is detected to be resolved within a preset time window, an emergency intervention command is generated and the emergency intervention command and related event data are pushed to the remote management terminal.

[0013] In a preferred embodiment of the present invention, the backend processing module further includes an extensible algorithm library for storing and updating different anomaly detection models. The anomaly detection models include a fatigue state recognition model based on deep learning and a blind zone intrusion detection model based on dynamic thresholds. The system supports remote updating and replacement of models in the algorithm library through the data transmission module.

[0014] The embodiments of the present invention bring the following beneficial effects: This invention provides an intelligent monitoring system for sanitation vehicles. A data transmission module is communicatively connected to a front-end sensing module to upload real-time data to a back-end processing module. The back-end processing module is also communicatively connected to the data transmission module to process and analyze the received real-time data, identifying abnormal states and generating control commands. An application execution module is communicatively connected to the back-end processing module to execute early warning prompts or vehicle control operations based on the control commands. Abnormal states include at least obstacles in the sanitation vehicle's blind spot and driver fatigue. This system integrates driver behavior monitoring and blind spot warnings to achieve real-time identification and tiered intervention of abnormal states, significantly improving the safety and management response efficiency of sanitation operations. Other features and advantages of this disclosure will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above.

[0015] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a structural diagram of an intelligent monitoring system for sanitation vehicles provided in an embodiment of the present invention; Figure 2 This is a structural diagram of another intelligent monitoring system for sanitation vehicles provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Vehicle-mounted monitoring systems are widely used in vehicle management, using devices such as GPS and cameras to collect vehicle location and video footage. However, existing systems are mostly limited to simple driving recording and positioning, lacking in-depth perception of driver status and proactive early warning capabilities for operational scenarios. Especially in sanitation vehicle operations, vehicles often face safety hazards such as large blind spots and driver fatigue. Existing technologies cannot identify and intervene in abnormal behaviors in real time and at different levels, resulting in delayed accident response and failing to meet the urgent needs of smart sanitation for safety and intelligent management.

[0020] Based on this, the present invention provides an intelligent monitoring system for sanitation vehicles. A data transmission module is communicatively connected to a front-end sensing module to upload real-time data to a back-end processing module. The back-end processing module is also communicatively connected to the data transmission module to process and analyze the received real-time data to identify abnormal states and generate control commands. An application execution module is communicatively connected to the back-end processing module to execute early warning prompts or vehicle control operations based on the control commands. Abnormal states include at least obstacles in the blind spot of the sanitation vehicle and driver fatigue. This approach, by integrating driver behavior monitoring and blind spot early warning, achieves real-time identification and tiered intervention of abnormal states, significantly improving the safety of sanitation operations and the efficiency of management response. To facilitate understanding of this embodiment, a detailed description of the intelligent monitoring system for sanitation vehicles disclosed in this embodiment of the invention will be provided first.

[0021] Example 1 This invention provides an intelligent monitoring system for sanitation vehicles. Figure 1 This is a structural diagram of an intelligent monitoring system for sanitation vehicles provided in an embodiment of the present invention. Figure 1 As shown, the intelligent monitoring system for sanitation vehicles can include the following structure: a front-end sensing module, a data transmission module, a back-end processing module, and an application execution module.

[0022] The front-end sensing module is used to collect real-time data from sanitation vehicles; the real-time data includes: operating status data, driver behavior data, and operating environment data.

[0023] The front-end sensing module refers to the collection of various sensors and data acquisition devices installed on the sanitation vehicle, responsible for acquiring raw data on vehicle operation, driver status, and the surrounding environment. Specifically, it includes devices such as GPS locators, speed sensors, DSM cameras, and blind spot radar. These devices are connected to the onboard terminal via the vehicle's CAN bus or a dedicated data cable for data aggregation.

[0024] The real-time data includes operational status data (such as vehicle speed, location, fuel consumption, and sweeping efficiency), driver behavior data (such as facial images and eyes-closed status), and working environment data (such as distance to obstacles in blind spots). The data collection frequency is configurable, typically set to 10 times per second to ensure real-time performance.

[0025] The data transmission module communicates with the front-end perception module and is used to upload real-time data to the back-end processing module.

[0026] The data transmission module is the communication unit responsible for sending data collected from the front end to the back end server. Specifically, it uses a 4G / 5G communication module or a WiFi module to establish a data transmission channel via a cellular network or local area network. In actual deployments, this module is usually integrated into the vehicle's intelligent terminal, supporting breakpoint resume functionality to ensure data is not lost when the signal is poor.

[0027] The backend processing module communicates with the data transmission module and is used to process and analyze the received real-time data to identify abnormal states and generate control commands.

[0028] The backend processing module refers to the software system deployed on cloud servers or local data centers. It receives data uploaded from the frontend, runs analysis algorithms, and identifies abnormal states. This module is deployed using a microservice architecture and includes sub-components such as data receiving services, algorithm analysis services, and alarm decision services, supporting high-concurrency data processing. The application execution module communicates with the backend processing module and is used to execute warning prompts or vehicle control operations based on control commands.

[0029] The application execution module refers to the terminal equipment that executes specific actions according to backend instructions, including vehicle alarms and vehicle control systems. Vehicle alarms are buzzers, LED indicator lights, or voice-activated horns installed in the driver's cab; the vehicle control system interacts with the vehicle's electronic control unit (ECU) via the CAN bus, sending control signals such as deceleration and stopping.

[0030] Abnormal conditions include, at a minimum, obstacles in the blind spot of the sanitation vehicle and driver fatigue.

[0031] Data interaction between modules is achieved via wired or wireless means. The front end and the data transmission module are connected via vehicle Ethernet or CAN bus; the data transmission module and the back end are connected via 4G / 5G mobile communication network; and the back end and the application execution module exchange commands via wireless network.

[0032] The interaction process between modules is as follows: the front-end perception module collects data → the data transmission module packages, encrypts, and uploads the data → the back-end processing module analyzes and identifies anomalies → the back-end generates control commands → the application execution module executes warnings or controls the vehicle.

[0033] The intelligent monitoring system for sanitation vehicles provided in this invention comprises a data transmission module communicatively connected to a front-end sensing module for uploading real-time data to a back-end processing module; a back-end processing module communicatively connected to the data transmission module for processing and analyzing the received real-time data to identify abnormal states and generate control commands; and an application execution module communicatively connected to the back-end processing module for executing early warning prompts or vehicle control operations based on the control commands. Abnormal states include at least obstacles in the blind spots of the sanitation vehicle's operation and driver fatigue. This approach, by integrating driver behavior monitoring and blind spot early warning, achieves real-time identification and tiered intervention of abnormal states, significantly improving the safety of sanitation operations and the efficiency of management response. Example 2 This invention also provides another intelligent monitoring system for sanitation vehicles; this intelligent monitoring system for sanitation vehicles is implemented based on the system described in the above embodiments.

[0034] Figure 2 A structural diagram of another intelligent monitoring system for sanitation vehicles provided in an embodiment of the present invention is shown below. Figure 2 As shown, the intelligent monitoring system for sanitation vehicles details the structural composition of each module: The front-end perception module includes: a vehicle status acquisition submodule, a driver status acquisition submodule, and an environmental perception submodule.

[0035] The vehicle status acquisition submodule is used to collect the driving and operational parameters of sanitation vehicles.

[0036] The vehicle status acquisition submodule consists of a GPS locator, speed sensor, and sweeping efficiency detector, which collects the sanitation vehicle's driving parameters (location, speed) and operational parameters (sweeping efficiency). The sweeping efficiency detector calculates the amount of garbage swept per unit area in real time using pressure and flow sensors installed on the sweeping brush or suction nozzle. The GPS locator, speed sensor, and sweeping efficiency detector are all communicatively connected to the data transmission module.

[0037] The GPS locator is used to obtain real-time vehicle latitude, longitude, speed, direction, altitude, and other information. It employs a BeiDou / GPS dual-mode positioning module, achieving meter-level positioning accuracy, supports AGPS assisted positioning, and has a cold start time of less than 30 seconds.

[0038] The speed sensor is usually a wheel speed sensor or a vehicle speed signal read from the CAN bus. When the GPS signal is weak (such as in a tunnel), the system automatically switches to the vehicle speed data read from the CAN bus to ensure data continuity.

[0039] The sweeping efficiency detector consists of a contact sensor installed on the sweeping brush and a material level sensor inside the waste bin. The contact sensor detects the contact pressure between the sweeping brush and the ground to determine the degree of brush wear; the material level sensor uses ultrasonic or laser ranging to monitor the waste bin's load in real time, prompting for emptying when the load reaches 80%.

[0040] The driver status acquisition submodule is used to collect the driver's facial features and behavioral posture.

[0041] The driver status acquisition submodule includes a DSM camera. The DSM camera is used to capture real-time facial images of the driver. It is mounted on the dashboard or A-pillar in front of the driver, with the lens pointed at the driver's face. The captured facial images are used to identify eye status, head posture, etc. This camera uses infrared illumination technology to ensure clear images in nighttime or backlit conditions, supports a wide dynamic range, and avoids interference from strong light.

[0042] The DSM camera features a resolution of ≥1080P, a frame rate of ≥30fps, and a built-in DSP processor, enabling local face detection and eye opening / closing recognition. An integrated AI algorithm chip allows for preliminary eye-closing detection at the front end, reducing data transmission volume.

[0043] The environmental perception submodule is used to collect information about obstacles around the sanitation vehicle.

[0044] The environmental perception submodule includes at least one blind spot monitoring radar or surround-view camera. This radar or camera is installed on the side or rear of the sanitation vehicle to detect blind spots around the vehicle and obtain detection data. The blind spot monitoring radar typically uses millimeter-wave radar with a detection range of 0.5-10 meters and can penetrate rain and fog. The surround-view camera provides 360-degree panoramic images for visualization assistance.

[0045] Among them, the blind spot monitoring radar adopts a 77GHz millimeter-wave radar with a detection angle of 120°. It can track multiple targets simultaneously and output the target distance, speed, and azimuth.

[0046] The surround-view camera uses four wide-angle cameras (front, rear, left, and right) to generate a panoramic top-down view through image stitching algorithms.

[0047] The data collected by each submodule of the front-end sensing module is aggregated by the data acquisition board of the vehicle terminal, and then transmitted to the data transmission module after preliminary formatting. The acquisition board has a built-in MCU, which is responsible for converting analog signals into digital signals and packaging them according to a unified protocol.

[0048] The backend processing module includes: a data analysis submodule and an alarm decision submodule.

[0049] The data analysis submodule is used to perform real-time analysis on the received real-time data based on a preset algorithm to determine whether there are any anomalies and obtain the analysis results.

[0050] The data analysis submodule is deployed on the algorithm engine of the backend server. It receives real-time data uploaded from the frontend and runs preset algorithms to detect anomalies. A stream processing framework (such as Apache Flink) is used to process the data stream in real time, with processing latency controlled within 100 milliseconds.

[0051] The alarm decision submodule, connected to the data analysis submodule, is used to generate tiered alarm commands based on the analysis results.

[0052] The alarm decision-making submodule generates alarm commands of different levels based on data analysis results and preset rules. The decision-making logic uses a rule engine (such as Drools), allowing managers to dynamically adjust alarm thresholds through a visual interface.

[0053] The data transmission module pushes data to the backend data access layer → the data analysis submodule subscribes to the data stream and performs analysis → the analysis results are pushed to the alarm decision submodule → the alarm decision submodule generates instructions according to the rules and sends them to the application execution module.

[0054] The backend processing module is also used to continuously receive feedback data from the front-end sensing module through the data transmission module after issuing the warning command. If the abnormal state is not detected to be resolved within the preset time window, an emergency intervention command is generated and the emergency intervention command and related event data are pushed to the remote management terminal.

[0055] The backend processing module continuously monitors and provides feedback after issuing an alert. If the anomaly is not resolved, it generates an emergency intervention command and pushes it to the remote management terminal.

[0056] The preset time window refers to the time interval from issuing an alert to reconfirming the abnormal state, which can be configured to 10 seconds or 15 seconds. Within the time window, the system continuously receives feedback data from the front-end sensing module.

[0057] If the abnormal state is not resolved after the emergency intervention window expires, the backend processing module will generate an intervention command of a higher level than the second-level alarm. This command includes: pushing an "Emergency Event" pop-up window to the remote management terminal; automatically dialing the preset emergency contact number; and forcibly sending a highest-priority parking command if the vehicle has not stopped.

[0058] Among them, the remote management terminal refers to the monitoring platform on the computer or mobile phone used by the administrator, which maintains a real-time connection with the backend via WebSocket, and pops up a red alarm window and plays a prompt sound when an emergency is received.

[0059] The backend issues a first or second level warning → starts a timer (e.g., 10 seconds) → continuously receives feedback data from the frontend → determines whether the anomaly has been resolved when the timer ends → if not, generates an emergency intervention command → sends it to the remote management terminal via a push service (e.g., JPush).

[0060] The backend processing module also includes an extensible algorithm library, which is used to store and update different anomaly detection models. The anomaly detection models include a fatigue state recognition model based on deep learning and a blind zone intrusion detection model based on dynamic thresholds. The system supports remote updating and replacement of models in the algorithm library through the data transmission module.

[0061] The scalable algorithm library is deployed in a model repository on the backend server, storing multiple anomaly detection models. It employs containerized deployment, with each model packaged as an independent Docker image, and uses Kubernetes for version management and elastic scaling.

[0062] The deep learning-based fatigue state recognition model employs lightweight convolutional neural networks such as ResNet or MobileNetV3, trained on an annotated driver face dataset. It can recognize various fatigue and distraction states, including closed eyes, yawning, head down, phone calls, and smoking. The model achieves an accuracy of ≥95% and an inference time of ≤50ms.

[0063] The blind spot intrusion detection model based on dynamic thresholds employs an adaptive threshold algorithm, dynamically adjusting the warning threshold according to vehicle speed, steering angle, and obstacle trajectory. For example, the warning distance is set to 3 meters when reversing at low speed and 5 meters when driving at high speed. The model predicts obstacle trajectories using Kalman filtering to anticipate collision risks in advance.

[0064] Remote updates and replacements include: the system supports OTA upgrades of models in the algorithm library via the data transmission module. The update process is as follows: the backend releases a new model version → the vehicle terminal detects the update notification → the new model is downloaded during non-operational periods (e.g., at night) → after verifying integrity, the old model is hot-loaded and replaced without requiring a system restart.

[0065] The data analysis submodule calls models from the algorithm library for inference at runtime. The models provide calling interfaces in the form of RESTful API or gRPC services, which facilitates dynamic calling by different analysis units.

[0066] The data analysis submodule includes: a fatigue driving analysis unit, a blind spot warning analysis unit, and a data fusion unit.

[0067] The fatigue driving analysis unit receives image data from the DSM camera and uses an image recognition algorithm to detect the duration of the driver's eye closure based on the image data.

[0068] The fatigue driving analysis unit receives image data from the DSM camera and detects the duration of eye closure using an image recognition algorithm. Specifically, the algorithm uses MTCNN for face detection, locates the eye region, and calculates the eyelid opening / closing degree (EAR value). Consecutive frames with EAR values ​​less than a threshold are considered as having closed eyes, and the duration of eye closure is timed and accumulated. Simultaneously, it detects yawning (mouth opening / closing degree) and head-down movements (head pitch angle) to comprehensively assess fatigue levels.

[0069] The blind spot early warning analysis unit is used to receive detection data from the environmental perception submodule and determine whether there are obstacles within the preset range of the sanitation vehicle.

[0070] The blind spot warning analysis unit receives detection data from radar or cameras to determine whether there are obstacles within a preset range. The preset range is configurable, typically set to 1.5 meters to the side and 3 meters behind as the warning zone; a warning is triggered when an obstacle enters the warning zone and the vehicle's turn signal is on or the vehicle speed exceeds 5 km / h.

[0071] The data fusion unit is used to perform correlation analysis between operational status data and driver behavior data in order to identify high-risk driving scenarios.

[0072] The data fusion unit analyzes the correlation between vehicle status data (vehicle speed, position, turn signal status) and driver behavior data. For example, when the vehicle speed is greater than 60 km / h and the driver closes their eyes for more than 1 second, it is identified as a "high-speed fatigue high-risk scenario"; when the vehicle turns right and the right-side radar detects a pedestrian, it is identified as a "turning blind spot high-risk scenario".

[0073] The three analysis units process different data sources in parallel, and the analysis results are aggregated into the data fusion unit for comprehensive scoring. The scoring results (such as fatigue risk level 0-100 points) are then transmitted to the alarm decision submodule.

[0074] The alarm decision submodule is configured with a first alarm threshold and a second alarm threshold.

[0075] The first alarm threshold is set as the boundary for judging a minor abnormal state. For example: the duration of closed eyes is ≥2 seconds but <4 seconds, or the distance of an obstacle in the blind spot is >1 meter. When this threshold is reached, the first-level alarm is triggered. The first-level alarm is a vehicle-mounted audio and visual warning: the buzzer sounds 3 times, a yellow warning light illuminates on the instrument panel, and a voice announcement says "Please pay attention to driving safety".

[0076] The second alarm threshold is set as the boundary for determining a severe abnormal state. For example: eyes closed for ≥4 seconds and the driver does not respond, or an obstacle within the blind spot is ≤0.5 meters away and the vehicle is still moving. Reaching this threshold triggers a second-level alarm. The second-level alarm includes: the backend server sending an emergency voice message "Please stop immediately" to the in-vehicle voice system via the 4G network; if there is still no response, a deceleration command is sent to the vehicle's ECU (reducing speed by 5 km / h every 2 seconds) until the vehicle comes to a complete stop and the hazard lights are activated.

[0077] When the data analysis submodule determines that the abnormal state has reached the first alarm threshold, the alarm decision submodule generates a first-level alarm command, which is used to trigger the audible and visual prompts of the vehicle terminal.

[0078] Specifically, when the data analysis submodule determines that the abnormal state continues and reaches the second alarm threshold, the alarm decision submodule generates a second-level alarm command, which is used to trigger remote voice intervention or forced vehicle control.

[0079] The alarm decision submodule generates different levels of instruction codes (such as Level1 and Level2) based on the type and severity of the anomaly. The instructions are encapsulated in JSON format and sent to the application execution module via the MQTT protocol.

[0080] The data transmission module includes a wireless communication unit and a data security unit.

[0081] The wireless communication unit uses 4G / 5G or WiFi communication technology to establish a real-time data link between the front-end sensing module and the back-end processing module.

[0082] It employs industrial-grade 4G / 5G modules (such as the Huawei ME909s), supports automatic multi-band switching, and has a built-in eSIM chip for easy remote carrier switching. In areas with no signal, the module automatically caches data (maximum cache 2GB) and resumes transmission after signal recovery.

[0083] The data security unit is responsible for data encryption and compression. Encryption uses the AES-256 algorithm with a key that is rotated periodically; compression uses the LZ4 algorithm, achieving a compression ratio of up to 50% and reducing bandwidth consumption. The security unit also includes a device authentication mechanism, with each vehicle having a unique device ID, and requiring two-way certificate authentication with the backend before data transmission.

[0084] The data security unit, connected to the wireless communication unit, is used to encrypt and compress the transmitted data.

[0085] The process involves: front-end data aggregation → encryption and compression by the security unit → establishment of a TCP / IP connection by the wireless communication unit → uploading to the back-end server via HTTPS or MQTT over TLS protocol.

[0086] The application execution module includes: vehicle alarm unit and vehicle control unit.

[0087] The vehicle-mounted alarm unit is installed in the cab of the sanitation vehicle and is used to issue visual or audible warnings based on the control of the back-end processing module.

[0088] The vehicle-mounted alarm unit, installed in the driver's cab, includes the following alarm devices: a buzzer (above 85dB); LED warning lights (red / yellow dual-color); and a voice broadcaster (TTS synthesized voice). The alarm unit receives alarm commands via the CAN bus and selects the corresponding alarm mode based on the command code.

[0089] The vehicle control unit communicates with the electronic control unit of the sanitation vehicle and is used to execute operations such as deceleration, activating hazard lights, or forced stopping based on control commands from the back-end processing module.

[0090] The controller that communicates with the vehicle control unit and the sanitation vehicle electronic control unit (ECU) is usually a separate VCU (vehicle controller) or a control module integrated into the vehicle terminal, which sends control frames to the ECU via the CAN bus.

[0091] The process involves the backend processing module issuing control commands, the data transmission module receiving and forwarding them to the vehicle terminal, and the vehicle terminal parsing the commands and sending corresponding signals to the alarm unit or control unit via the CAN bus. When the control unit performs deceleration, it first sends a throttle limit signal, followed by a braking request signal to ensure smooth deceleration.

[0092] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0093] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0094] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent monitoring system for sanitation vehicles, characterized in that, The system includes: a front-end perception module, a data transmission module, a back-end processing module, and an application execution module; The front-end sensing module is used to collect real-time data from the sanitation vehicle; the real-time data includes: operating status data, driver behavior data, and operating environment data. The data transmission module is communicatively connected to the front-end sensing module and is used to upload the real-time data to the back-end processing module. The back-end processing module is communicatively connected to the data transmission module and is used to process and analyze the received real-time data in order to identify abnormal states and generate control commands. The application execution module is communicatively connected to the backend processing module and is used to execute early warning prompts or vehicle control operations based on the control commands. The abnormal conditions include at least the blind spots of sanitation vehicles and driver fatigue.

2. The intelligent monitoring system for sanitation vehicles according to claim 1, characterized in that, The front-end perception module includes: a vehicle status acquisition submodule, a driver status acquisition submodule, and an environmental perception submodule; The vehicle status acquisition submodule is used to collect the driving parameters and operating parameters of the sanitation vehicle. The driver status acquisition submodule is used to acquire the driver's facial features and behavioral postures; The environmental perception submodule is used to collect information about obstacles around the sanitation vehicle.

3. The intelligent monitoring system for sanitation vehicles according to claim 2, characterized in that, The vehicle status acquisition submodule includes: a GPS locator, a speed sensor, and a cleaning efficiency detector; the GPS locator, the speed sensor, and the cleaning efficiency detector are all communicatively connected to the data transmission module; The driver status acquisition submodule includes: a DSM camera; the DSM camera is used to capture the driver's facial image in real time; The environmental perception submodule includes at least one blind spot monitoring radar or surround view camera; the blind spot monitoring radar or surround view camera is installed on the side or rear of the sanitation vehicle to detect visual blind spots around the vehicle body and obtain detection data.

4. The intelligent monitoring system for sanitation vehicles according to claim 1, characterized in that, The backend processing module includes: a data analysis submodule and an alarm decision submodule; The data analysis submodule is used to perform real-time analysis on the received real-time data based on a preset algorithm to determine whether there are any anomalies and obtain analysis results. The alarm decision submodule is connected to the data analysis submodule and is used to generate graded alarm commands based on the analysis results.

5. The intelligent monitoring system for sanitation vehicles according to claim 4, characterized in that, The data analysis submodule includes: a fatigue driving analysis unit, a blind spot warning analysis unit, and a data fusion unit; The fatigue driving analysis unit is used to receive image data from the DSM camera and detect the duration of the driver's eye closure based on the image data using an image recognition algorithm. The blind spot early warning analysis unit is used to receive the detection data from the environmental perception submodule and determine whether there are obstacles within the preset range of the sanitation vehicle; The data fusion unit is used to perform correlation analysis between the operating status data and the driver behavior data in order to identify high-risk driving scenarios.

6. The intelligent monitoring system for sanitation vehicles according to claim 4, characterized in that, The alarm decision submodule is configured with a first alarm threshold and a second alarm threshold, wherein: When the data analysis submodule determines that the abnormal state has reached the first alarm threshold, the alarm decision submodule generates a first-level alarm instruction, which is used to trigger the audible and visual prompts of the vehicle terminal. When the data analysis submodule determines that the abnormal state continues and reaches the second alarm threshold, the alarm decision submodule generates a second-level alarm command, which is used to trigger remote voice intervention or forced vehicle control.

7. The intelligent monitoring system for sanitation vehicles according to claim 1, characterized in that, The data transmission module includes: a wireless communication unit and a data security unit; The wireless communication unit uses 4G / 5G or WiFi communication technology to establish a real-time data link between the front-end sensing module and the back-end processing module. The data security unit is connected to the wireless communication unit and is used to encrypt and compress the transmitted data.

8. The intelligent monitoring system for sanitation vehicles according to claim 1, characterized in that, The application execution module includes: an in-vehicle alarm unit and a vehicle control unit; The vehicle alarm unit is installed in the driver's cab of the sanitation vehicle and is used to issue visual or audible warnings based on the control of the back-end processing module. The vehicle control unit is communicatively connected to the electronic control unit of the sanitation vehicle and is used to perform operations such as deceleration, activating hazard lights, or forced stopping based on the control commands of the back-end processing module.

9. The intelligent monitoring system for sanitation vehicles according to claim 1, characterized in that, The backend processing module is also used to continuously receive feedback data from the frontend sensing module through the data transmission module after issuing the warning command. If no abnormal state is detected to be resolved within a preset time window, an emergency intervention command is generated and the emergency intervention command and related event data are pushed to the remote management terminal.

10. The intelligent monitoring system for sanitation vehicles according to claim 1, characterized in that, The backend processing module also includes an extensible algorithm library for storing and updating different anomaly detection models. The anomaly detection models include a fatigue state recognition model based on deep learning and a blind zone intrusion detection model based on dynamic thresholds. The system supports remote updating and replacement of the models in the algorithm library through the data transmission module.

Images