Intelligent environmental sanitation emergency cleaning system based on multi-end cooperation

Abstract

This invention discloses an intelligent sanitation emergency cleaning system based on multi-terminal collaboration, comprising: an intelligent trash can, used to send an overflow alarm signal to a cloud-based dispatch platform when the trash can's filling status data reaches an overflow threshold; an intelligent wearable device, used to collect the location coordinates of workers in real time and upload them to the cloud-based dispatch platform, and to receive composite task dispatch instructions; the cloud-based dispatch platform, used to lock the corresponding intelligent trash can as a target trash can in response to the received overflow alarm signal; to match the target workers within the response range based on the location coordinates uploaded by the intelligent wearable device; to generate an emergency cleaning route based on the location coordinates of the target workers and the location coordinates of the target trash can; and to generate a composite task dispatch instruction including the target trash can and associated nodes in the cleaning queue, and send it to the intelligent wearable device of the target workers.

Description

Technical Field

[0001] This invention belongs to the field of smart city construction technology, specifically relating to an intelligent sanitation emergency cleaning system based on multi-terminal collaboration. Background Technology

[0002] With the advancement of smart city construction, urban sanitation management is gradually transforming from the traditional fixed-time and fixed-location collection model to digitalization and intelligence. Traditional sanitation operations are usually carried out according to fixed routes and schedules. This model cannot cope with sudden overflows of garbage bins caused by fluctuations in pedestrian traffic, which can easily lead to garbage accumulation and affect the city's appearance and environmental sanitation.

[0003] Existing smart sanitation solutions typically employ IoT technology, installing sensors inside trash cans to monitor their fullness. When a trash can is detected to be overflowing, the system generates an alarm signal and dispatches vehicles or personnel to clean it. However, most existing dispatch systems use a point-to-point, single-response model: an overflow point is detected, a person is dispatched to clean it, and the task ends or the person returns after completion. Existing dispatch logic often overlooks the possibility of other trash cans that are about to overflow or are conveniently located along the path of the worker heading to the target overflow point. This can lead to workers passing by a trash can that is 80% full, cleaning it, and leaving, only to have that same trash can soon overflow and trigger an alarm, requiring repeated trips and increasing empty runs and energy consumption. Traditional dispatch information is mostly pushed via mobile apps, which workers cannot easily view in real time while driving or working, making it easy to miss urgent tasks; furthermore, relying solely on manual map navigation makes it difficult to achieve a second-level response in the event of a sudden overflow.

[0004] Therefore, there is an urgent need for a multi-terminal collaborative system that can combine real-time location, dynamic path planning, and smart wearable devices to solve the above problems. Summary of the Invention

[0005] In view of this, the main objective of the present invention is to provide an intelligent sanitation emergency cleaning system based on multi-terminal collaboration.

[0006] To achieve the above objectives, the technical solution of the present invention is implemented as follows: A multi-terminal collaborative intelligent sanitation emergency cleaning system includes: several intelligent trash cans distributed in urban areas, several intelligent wearable devices worn by workers, and a cloud-based dispatch platform that communicates with the intelligent trash cans and the intelligent wearable devices. The smart trash can is equipped with a filling monitoring sensor and a first communication module, which is used to monitor the filling status data inside the can in real time, and send an overflow alarm signal to the cloud scheduling platform when the filling status data reaches the overflow threshold. The smart wearable device is equipped with a positioning module and a second communication module, which are used to collect the location coordinates of the workers in real time and upload them to the cloud dispatch platform, as well as to receive and display instructions from the cloud dispatch platform in a tactile or visual manner. The cloud-based scheduling platform is used to respond to the received overflow alarm signal by locking the corresponding smart trash can as the target trash can; matching the target worker within the response range based on the location coordinates uploaded by the smart wearable device; generating an emergency cleaning route based on the location coordinates of the target worker and the target trash can; establishing a spatial buffer zone of a preset width with the emergency cleaning route as the axis, and retrieving other smart trash cans located within the spatial buffer zone as associated nodes; obtaining the current filling status data of the associated nodes, and adding the associated nodes to the cleaning queue if the preset route-following cleaning conditions are met; generating a composite task scheduling instruction containing the target trash can and the associated nodes in the cleaning queue, and sending it to the smart wearable device of the target worker.

[0007] Preferably, the cloud-based dispatch platform is specifically used to acquire traffic topology data of the roads traversed by the emergency clearing route, identify the direction of travel and median strip information; based on the traffic topology data, eliminate areas that are physically adjacent to the emergency clearing route but cannot be legally entered by changing lanes directly; determine a set of areas that can be reached and returned to the original route from any node on the emergency clearing route within a preset additional travel time threshold, and construct the set of areas as the spatial buffer zone; the cloud-based dispatch platform is also used to acquire dynamic traffic data related to the emergency clearing route in real time, and dynamically scale the preset width of the spatial buffer zone and the additional travel time threshold according to the road congestion index; wherein, the conditions for clearing along the route include the road lateral direction of the associated node being consistent with the current travel direction of the workers, or the existence of a U-turn intersection at a distance less than a preset value.

[0008] Preferably, the cloud-based dispatch platform is specifically used to, when matching the target workers within the response range, also receive the current load status data of the vehicle driven by the worker uploaded by the smart wearable device; analyze the historical average compression density contained in the overflow alarm signal of the target trash can to estimate the weight of the target trash; calculate the remaining load capacity of the vehicle, and only when the remaining load capacity is greater than the sum of the weight of the target trash and the preset safety redundancy, mark the corresponding worker as a candidate target worker; if all workers within the response range do not meet the capacity verification step, then generate a relay cleaning instruction and dispatch empty vehicles to provide support.

[0009] Preferably, the smart wearable device is equipped with a near-field communication module; The smart trash can is equipped with a corresponding electronic tag recognition module; The smart wearable device is also used to perform near-field handshake communication with the electronic tag recognition module of the smart trash can through the near-field communication module when the target worker arrives at the target trash can or the associated node location for cleaning, and to generate a human-trash can interaction verification code. The smart trash can is also used to generate an emptying reset signal after it is determined to be emptied by monitoring sensors; The cloud-based scheduling platform is also used to determine that the cleaning task of the smart trash can has been completed when it simultaneously receives an emptying and reset signal from the smart trash can and a human-trash can interaction verification code generated from near-field handshake communication, and automatically removes the node corresponding to the smart trash can from the navigation queue on the smart wearable device.

[0010] Preferably, the filling monitoring sensor of the smart trash can includes a multispectral camera component and a visible light camera component; The cloud-based scheduling platform is also used to identify surface coverings in images captured by the visible light camera component before generating composite task scheduling instructions, and to determine whether there are any obstructions that hinder spectral penetration; if there are no obstructions, spectral analysis is performed on the garbage images in the target garbage bin and the associated node to identify the hazard level and recycling category of the garbage content; the loading type attribute of the vehicle equipped by the target operator is obtained; if the garbage recycling category of the associated node does not match the loading type attribute, even if the associated node is located within the spatial buffer zone, it is prohibited from being added to the cleaning queue, and a sorting and transfer task is generated.

[0011] Preferably, the smart trash can is equipped with an adaptive frequency adjustment module, which is also used to monitor the rate of change of filling status data in real time; when the rate of change exceeds a preset sudden growth threshold, the sampling frequency of the first communication module sending data to the cloud scheduling platform is increased; when the filling status data is detected to remain unchanged for a preset duration, it enters a low-power sleep mode and sends status data only at a preset heartbeat time each day; the smart trash can also includes a power monitoring module, and the cloud scheduling platform dynamically adjusts the upper limit of the sampling frequency increase based on the remaining power fed back by the power monitoring module; The cloud-based scheduling platform is also used to predict the remaining time window before the smart trash can reaches the overflow state based on the rate of change, and to issue the overflow alarm signal in advance before the remaining time window is less than the time required for the operator to arrive.

[0012] Preferably, the cloud-based dispatching platform further includes a heat map generation module for sudden events, which is also used to statistically analyze the spatiotemporal distribution of overflow alarm signals issued by smart trash cans in each area within a preset period; identify hotspot areas with high frequency of overflow; and send pre-dispatch instructions to the smart wearable devices of idle workers located at the edge of the hotspot area during the time period corresponding to the hotspot area, so as to guide the workers to move towards the center of the hotspot area for patrol, thereby shortening the physical distance for responding to sudden overflow events.

[0013] Preferably, the smart trash can is also equipped with an odor gas sensor, which is also used to upload odor concentration data; The cloud-based scheduling platform is also used to determine the urgency of the cleanup. , where α and β are weighting coefficients, the regional background reference value is uploaded in real time by environmental monitoring nodes deployed in the same area, and is used to filter out environmental background odor noise; when the cleaning urgency result S exceeds the preset alarm baseline, even if the filling rate has not reached the physical overflow state, the overflow alarm signal is forcibly triggered and marked as a high-priority deodorization task in the generated composite task scheduling instruction.

[0014] Preferably, the smart wearable device is a smart helmet or AR glasses; The smart wearable device is also used to overlay and project the emergency cleaning route and the location of the associated nodes onto the real-world scene on the lens display of the smart helmet or AR glasses. The smart wearable device uses a fusion positioning algorithm, combined with visual odometry and inertial measurement unit, to compensate and correct the projection position of the emergency cleaning route using inertial navigation data when environmental visual features are lost or GNSS signals are weak. When the operator approaches the target trash can or associated node, the current filling percentage and internal trash type of the trash can are displayed intuitively on the display in the form of augmented reality (AR) tags, assisting the operator in preparing the corresponding cleaning tools in advance.

[0015] Preferably, after sending the composite task scheduling instruction, the cloud scheduling platform continuously receives the location coordinates of the smart wearable device; when it detects that the real-time movement trajectory of the target worker continuously deviates from the emergency cleanup route by more than the allowable threshold, or that the dwell time at an associated node is abnormal, it sends an alarm to the administrator, sends a reminder inquiry to the smart wearable device, or initiates a backup worker matching process.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention generates emergency cleaning routes through a cloud-based scheduling platform and establishes a spatial buffer zone around these routes. It not only handles overflowing target trash cans but also automatically retrieves other trash cans along the route that meet the road cleaning conditions, packages them into composite task scheduling instructions, and transforms a single emergency task into a route area task. With a single dispatch, it simultaneously solves the problems of sudden overflow points and potential overflow risk points along the route, avoids repeated trips by operators on the same road segment, greatly reduces vehicle empty running rate, and improves the output ratio of a single operation. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the structure of an intelligent sanitation emergency cleaning system based on multi-terminal collaboration, provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] In the accompanying drawings of this embodiment, the same or similar reference numerals correspond to the same or similar components. In the description of this invention, it should be understood that the terms "upper," "lower," "left," "right," "inner," "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the accompanying drawings are only for illustrative purposes and should not be construed as limiting this patent. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0020] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, article, or apparatus that includes that element.

[0021] This invention provides an intelligent sanitation emergency cleaning system based on multi-terminal collaboration, such as...Figure 1 As shown, the system includes: several smart trash cans distributed in urban areas, several smart wearable devices worn by operators, and a cloud-based dispatch platform that communicates with the smart trash cans and the smart wearable devices; The smart trash can is equipped with a filling monitoring sensor and a first communication module, which is used to monitor the filling status data inside the can in real time, and send an overflow alarm signal to the cloud scheduling platform when the filling status data reaches the overflow threshold. The smart wearable device is equipped with a positioning module and a second communication module, which are used to collect the location coordinates of the workers in real time and upload them to the cloud dispatch platform, as well as to receive and display instructions from the cloud dispatch platform in a tactile or visual manner. The cloud-based scheduling platform is used to respond to the received overflow alarm signal by locking the corresponding smart trash can as the target trash can; matching the target worker within the response range based on the location coordinates uploaded by the smart wearable device; generating an emergency cleaning route based on the location coordinates of the target worker and the target trash can; establishing a spatial buffer zone of a preset width with the emergency cleaning route as the axis, and retrieving other smart trash cans located within the spatial buffer zone as associated nodes; obtaining the current filling status data of the associated nodes, and adding the associated nodes to the cleaning queue if the preset route-following cleaning conditions are met; generating a composite task scheduling instruction containing the target trash can and the associated nodes in the cleaning queue, and sending it to the smart wearable device of the target worker.

[0022] This invention generates emergency cleanup routes through a cloud-based scheduling platform and establishes spatial buffer zones centered on these routes. It not only handles overflowing target trash cans but also automatically retrieves other trash cans along the route that meet the road cleanup requirements, packaging them into composite task scheduling instructions, transforming a single emergency task into a route-area task. With a single deployment, it simultaneously addresses both sudden overflow points and potential overflow risks along the route, avoiding repeated trips by personnel on the same road segment, significantly reducing vehicle empty runs, and improving the output ratio of a single operation.

[0023] This invention introduces a smart wearable device worn by workers as an interactive terminal. Its positioning module uploads coordinates in real time and displays instructions through touch or vision. Compared to traditional mobile app notifications, smart wearable devices (such as vibration bracelets or glasses displays) allow workers to perceive emergency tasks immediately, even when their hands are occupied or while driving, without needing to unlock their phones. This ensures that after an overflow alarm signal is issued, the system can quickly match personnel within range and directly reach their sensing terminals, truly achieving rapid response to emergencies.

[0024] When generating tasks, this invention not only identifies overflowing bins, but also retrieves associated nodes within the spatial buffer zone to include bins that are not yet completely overflowing but are along the route into the cleaning queue. By cleaning up bins that are about to overflow along the route, it proactively eliminates future alarm sources, reduces the frequency of alarms and the number of emergencies in subsequent periods, thereby optimizing the allocation of sanitation resources in the entire jurisdiction from a macro perspective and extending the effective service time of the bins.

[0025] In some embodiments, the cloud-based dispatch platform is specifically used to acquire traffic topology data of the roads traversed by the emergency clearing route, identify the direction of travel and median strip information of the roads; based on the traffic topology data, eliminate areas that are physically adjacent to the emergency clearing route but cannot be legally changed lanes to enter directly; determine a set of areas that can be reached and returned to the original route from any node on the emergency clearing route within a preset additional travel time threshold, and construct the set of areas as the spatial buffer zone; the cloud-based dispatch platform is also used to acquire dynamic traffic data related to the emergency clearing route in real time, and dynamically scale the preset width of the spatial buffer zone and the additional travel time threshold according to the road congestion index; wherein, the conditions for clearing along the route include the road lateral direction of the associated node being consistent with the current travel direction of the workers, or the existence of a U-turn intersection at a distance less than a preset value.

[0026] In this embodiment, when constructing the spatial buffer zone, the cloud-based scheduling platform abandons the traditional coarse-grained division method based solely on geometric distance (such as a 500-meter circle radiating outward from the route center). Instead, it adopts accessibility analysis based on the road network topology. Specifically, it retrieves high-precision map data or traffic network APIs to obtain detailed attributes of the road segments covered by the emergency clearing route, including but not limited to lane direction (one-way / two-way), the presence of physical barriers (such as green belts or median barriers), and no-lane-change signs. Based on this data, when a region is detected that is geographically very close to the current route but has an insurmountable physical barrier or solid line prohibition, the region is determined to be visually adjacent but logically unreachable and is directly removed from the buffer zone candidate, thereby preventing dispatchers from making illegal lane changes or unnecessary detours.

[0027] Furthermore, the cloud-based dispatch platform integrates with third-party traffic network APIs to obtain real-time dynamic traffic data related to the emergency cleanup routes. The platform dynamically adjusts the preset width of the spatial buffer zone and the additional travel time threshold based on the road segment's congestion index (e.g., a quantitative indicator calculated from real-time vehicle speed, queue length, or delay ratio). For example, when a target road segment is detected to be severely congested (e.g., the traffic congestion index exceeds a preset alarm threshold), the system automatically narrows the spatial buffer zone radius and reduces the allowed additional travel time threshold, thereby prioritizing the ability of personnel to concentrate resources to complete the emergency cleanup of the target trash cans and effectively preventing the timeliness of core emergency tasks from being compromised due to the cleanup of related nodes along the route.

[0028] Furthermore, the generation logic of the spatial buffer zone is based on time cost rather than "spatial radius." Using a series of discrete nodes on the generated emergency cleanup route as starting anchor points, a path search algorithm (such as an improved Dijkstra's algorithm or an isochronous circle algorithm) is run to simulate the complete process of workers leaving the current route, reaching a candidate area for cleanup, and then returning to the original route or heading to the next target point. The additional travel time generated by this process is calculated, and only the set of areas where the additional time consumption is less than a preset threshold (which is dynamically adjusted according to road conditions) is included in the spatial buffer zone. This ensures that the task of incidental cleanup is carried out without significantly delaying the main emergency task, achieving a balance between emergency response speed and overall operational efficiency.

[0029] Furthermore, for associated nodes located within the spatial buffer zone, vector analysis is used to compare the current travel direction vector of the operator with the lateral attributes of the road where the associated node is located. If the associated node is located on the right side of the road and is in the same direction as the vehicle's current travel, it is determined to be completely on the same route; if the associated node is located on the opposite side of the road (left side), a further search is conducted to determine if there is a legal U-turn or left-turn intersection within a preset distance (e.g., 200 meters) ahead or behind. Only when a legal U-turn path exists and the detour cost of that path is within the allowable range is the condition of being on the same route met; otherwise, even if the node is very close, it will be ignored to prevent the operator from performing dangerous and illegal operations in busy road sections.

[0030] In some embodiments, the cloud-based dispatching platform is specifically used to, when matching the target workers within the response range, also receive the current load status data of the vehicle driven by the worker uploaded by the smart wearable device; analyze the historical average compression density contained in the overflow alarm signal of the target trash can to estimate the weight of the target trash; calculate the remaining load capacity of the vehicle, and only when the remaining load capacity is greater than the sum of the weight of the target trash and the preset safety redundancy, mark the corresponding worker as a candidate target worker; if all workers within the response range do not meet the capacity verification step, generate a relay cleaning instruction and dispatch an empty vehicle to provide support.

[0031] In this preferred embodiment of personnel matching on the cloud-based dispatch platform, the sanitation vehicles (such as small collection vehicles or compactors) driven by the operators are equipped with onboard weighing sensors (such as axle load meters or suspension weighing modules). This data is synchronized in real time to the smart wearable devices worn by the operators via Bluetooth or vehicle bus, and then uploaded to the cloud-based dispatch platform by the smart wearable devices. After identifying several candidate personnel within the response range, the cloud platform first reads the current load status data of their vehicles (e.g., current loaded weight or remaining volume percentage), thereby accurately grasping the real-time throughput capacity of each operating vehicle.

[0032] Meanwhile, for the target trash can that triggered the alarm, due to the significant differences in trash density between different areas (such as the food court and office area), the cloud-based dispatch platform will retrieve the historical collection records of the target trash can and analyze its historical average compression density. The cloud-based dispatch platform will then multiply the fixed volume of the trash can, the current filling percentage, and the historical average compression density to calculate the estimated weight of the trash to be collected.

[0033] To prevent exceeding the weight limit due to sensor errors or rain-induced water absorption by the garbage, a preset safety margin (e.g., 10%-15% of the estimated value) is added to the estimated weight. Then, the vehicle's remaining load capacity is compared to the sum of the estimated weight and the safety margin. Only those personnel whose remaining load capacity meets the requirements are retained in the candidate list, effectively avoiding ineffective scheduling and wasted empty runs caused by personnel arriving but not being able to fit on the vehicle.

[0034] Furthermore, after traversing and searching within the response range, if the cloud-based dispatch platform finds that the remaining load of all nearby workers' vehicles is insufficient to accommodate the estimated weight of the target garbage bin (i.e., all workers do not meet the capacity verification steps), it will no longer forcibly assign nearby patrol vehicles. Instead, it will generate a high-level relay cleaning instruction and directly dispatch the nearest large transfer vehicle that is empty or underloaded to provide support; or send a reservation instruction to the fully loaded workers to unload and return, ensuring that the sudden overflow event is handled with definite and physically feasible measures, rather than just a formality of dispatching orders.

[0035] In some embodiments, the smart wearable device is equipped with a near-field communication module; The smart trash can is equipped with a corresponding electronic tag recognition module; The smart wearable device is also used to perform near-field handshake communication with the electronic tag recognition module of the smart trash can through the near-field communication module when the target worker arrives at the target trash can or the associated node location for cleaning, and to generate a human-trash can interaction verification code. The smart trash can is also used to generate an emptying reset signal after it is determined to be emptied by monitoring sensors; The cloud-based scheduling platform is also used to determine that the cleaning task of the smart trash can has been completed when it simultaneously receives an emptying and reset signal from the smart trash can and a human-trash can interaction verification code generated from near-field handshake communication, and automatically removes the node corresponding to the smart trash can from the navigation queue on the smart wearable device.

[0036] In this embodiment, when a target worker wearing a smart wearable device (such as a smart bracelet with a built-in NFC chip) enters the preset near-field sensing range (e.g., 0-10 cm) of the target trash can, the active near-field communication module of the smart wearable device will activate the electronic tag identification module on the trash can or wake up the low-power Bluetooth beacon. Both parties then initiate an encrypted handshake protocol. The smart trash can sends its unique device ID and a current timestamp random number. After receiving this, the smart wearable device, combined with its own worker ID, uses a hash algorithm (such as SHA-256) to generate a time-sensitive human-trash interaction verification code. This verification code not only proves that the worker has indeed physically arrived at the designated location, but also, due to the inclusion of a timestamp and a random factor, effectively prevents cheating through replay attacks by recording signals, ensuring the authenticity of the attendance data.

[0037] Meanwhile, the filling monitoring sensors (such as ultrasonic probes or infrared beams) of the smart trash can continuously monitor the state inside the can. When the monitored value suddenly drops from a high level (overflowing state) to a low level (empty state, such as filling rate <5%), a signal is not triggered immediately. Instead, a state stabilization timer is started (for example, for 5 seconds). Only when the filling data remains stable at a low level after the timer ends, and the net weight displayed by the weight sensor returns to zero, is a valid emptying operation determined to be completed, and an emptying and reset signal is generated. This effectively filters out false alarms caused by trash can shaking, falling objects, or momentary blind spots of the sensors, ensuring the accuracy of the operation results.

[0038] The cloud-based scheduling platform has an allowed time deviation window (e.g., 3 minutes). The judgment condition is met only when, within the same time window, both the human-to-trash interaction verification code from the smart wearable device and the emptying / reset signal from the smart trash can are received simultaneously. At this point, the cloud-based scheduling platform confirms the cleaning task is complete, immediately sends an instruction to the smart wearable device to update the task status list, automatically removes the navigation node corresponding to that trash can from the current job queue, and plans the path to the next node. If the platform only receives the verification code without a reset signal (human present but no work performed), or only receives the reset signal without a verification code (due to unassigned personnel cleaning or sensor malfunction), the task will not be cancelled, and an exception work order will be triggered for administrator review.

[0039] In some embodiments, the filling monitoring sensor of the smart trash can includes a multispectral camera component and a visible light camera component; The cloud-based scheduling platform is also used to identify surface coverings in images captured by the visible light camera component before generating composite task scheduling instructions, and to determine whether there are any obstructions that hinder spectral penetration; if there are no obstructions, spectral analysis is performed on the garbage images in the target garbage bin and the associated node to identify the hazard level and recycling category of the garbage content; the loading type attribute of the vehicle equipped by the target operator is obtained; if the garbage recycling category of the associated node does not match the loading type attribute, even if the associated node is located within the spatial buffer zone, it is prohibited from being added to the cleaning queue, and a sorting and transfer task is generated.

[0040] In this embodiment, the multispectral camera component not only acquires visible light images but also covers specific bands such as near-infrared (NIR). When the trash can detects disposal behavior or is activated at a set time, it performs a multi-band scan of the surface trash inside the can. After receiving the raw spectral data, the cloud scheduling platform first performs radiometric calibration and atmospheric correction preprocessing, and then uses a spectral angle mapping algorithm to analyze the material composition. A library of standard spectral reflectance curves for common wastes (such as PET plastic, aluminum cans, organic kitchen waste, lithium batteries, etc.) is pre-set. By calculating the cosine of the angle between the acquired pixel spectral vector and the standard reference vector in high-dimensional space, it is possible to penetrate the stains on the surface of the trash and accurately identify the material properties of the trash, thereby determining its recycling category (such as recyclables, kitchen waste) and potential hazard level (such as identifying the characteristic spectrum of batteries or chemical containers).

[0041] Based on the above identification results, the cloud-based dispatch platform maintains a dynamic vehicle loading attribute matrix, which records in detail the legal loading type of the vehicle currently driven by the operator (e.g., food waste sealing vehicle only, general compactor vehicle, or hazardous waste transport vehicle). Before generating task instructions, the main waste category identified by the associated nodes is compared with the vehicle attributes of the target operator.

[0042] If the comparison results show a mismatch (for example, the operator is driving a special vehicle for food waste, while the associated nodes along the route are identified as primarily filled with cardboard boxes and foam plastic), even if the associated node is geographically located entirely within the spatial buffer zone and along the route, the classification isolation interlocking mechanism will be forcibly triggered. This mechanism, on the one hand, blocks the node in the current operator's task list to prevent illegal mixed loading and transportation; on the other hand, it automatically generates an independent classification and transfer task and assigns it to the allocation pool, waiting for subsequent dispatch of other operators with the corresponding recycling qualifications (such as recyclable waste transport vehicles) for specialized cleanup, thereby improving efficiency while strictly adhering to environmental regulations on waste classification.

[0043] In some embodiments, the smart trash can is equipped with an adaptive frequency adjustment module, which is also used to monitor the rate of change of filling status data in real time; when the rate of change exceeds a preset burst growth threshold, the sampling frequency of the first communication module sending data to the cloud scheduling platform is increased; when the filling status data is detected to remain unchanged for a preset duration, a low-power sleep mode is entered, and status data is sent only at a preset heartbeat time each day; the smart trash can also includes a power monitoring module, and the cloud scheduling platform dynamically adjusts the upper limit of the sampling frequency increase based on the remaining power fed back by the power monitoring module; The cloud-based scheduling platform is also used to predict the remaining time window before the smart trash can reaches the overflow state based on the rate of change, and to issue the overflow alarm signal in advance before the remaining time window is less than the time required for the operator to arrive.

[0044] In this embodiment, the smart trash can does not use a fixed-period communication mode, but is configured with "event-driven adaptive frequency adjustment logic. It calculates the first derivative (i.e., filling rate) of the filling status data in real time. When it is in certain sudden periods (such as lunchtime peak or during large events), if the filling rate ΔV / Δt exceeds the preset sudden growth threshold, it immediately enters a high-frequency agile mode, automatically compressing the interval between data sampling and uploading from the usual (e.g., once every 15 minutes) to the second or minute level (e.g., once every 30 seconds), ensuring that the cloud can capture the accurate moment when the trash can is full, avoiding the lag phenomenon of "actually full but the system shows not full" caused by data delay. Conversely, at night or during periods of low traffic, if the fluctuation amplitude of the filling data is detected to be lower than the noise threshold within a continuous preset duration (e.g., 2 hours), it automatically switches to an ultra-low power sleep mode, shutting down the high-energy-consuming communication module and retaining only a minimum of timers to send a device survival signal once a day at a preset heartbeat time (e.g., 3 am), thereby greatly extending the device's battery life.

[0045] In addition, the smart trash can is equipped with a power monitoring module. The cloud-based scheduling platform obtains the remaining power from the power monitoring module in real time and dynamically adjusts the upper limit of the sampling frequency based on the remaining power. Specifically, the cloud platform has preset multi-level energy management strategies: when the power is sufficient, the sampling frequency is allowed to be increased to the maximum value to ensure the highest sensitivity; when the remaining power is detected to be lower than a first preset threshold (e.g., 30%), the cloud platform will limit the increase in the sampling frequency; if the power further drops to a second preset threshold (e.g., 15%), even if the filling rate is extremely fast, the system will lock the sampling frequency at a low energy consumption level to ensure that the device will not completely lose connection due to excessive consumption before being charged or having its battery replaced.

[0046] In conjunction with the aforementioned front-end mechanism, the cloud-based scheduling platform no longer simply triggers an alarm based on whether the current filling rate has reached 90%. Instead, it combines the current filling rate with historical growth curves through linear regression or multinomial fitting to calculate the estimated remaining time window before the trash can reaches 100% overflow. For example, even if the current filling rate is only 70%, if an extremely rapid filling rate is detected, predicting that it will overflow in just 20 minutes, this situation is considered more urgent than a trash can with an 85% filling rate but currently stationary. Based on this prediction, the cloud-based scheduling platform obtains in real-time the estimated arrival time required for the target worker to travel from their current location to the trash can. The platform compares the estimated remaining time window with the estimated arrival time in real time. Once the trigger condition of "estimated remaining time window ≤ estimated arrival time + buffer time δ" is met, that is, the predicted overflow countdown is about to be less than the travel time, the platform will immediately issue an overflow alarm signal in advance. This ensures that when the operator arrives at the site at the estimated arrival time, the smart trash can is just about to overflow. This achieves a seamless connection between vehicle arrival and full bin, preventing garbage overflow and environmental pollution, and eliminating the waste of transportation capacity caused by premature cleaning.

[0047] In some embodiments, the cloud-based scheduling platform further includes a heat map generation module for sudden events, which is also used to statistically analyze the spatiotemporal distribution of overflow alarm signals issued by smart trash cans in each area within a preset period; identify hotspot areas with high frequency of overflow; and send pre-scheduling instructions to the smart wearable devices of idle workers located at the edge of the hotspot area during the time period corresponding to the hotspot area, so as to guide the workers to move towards the center of the hotspot area for patrol, thereby shortening the physical distance for responding to sudden overflow events.

[0048] In this embodiment, the cloud-based scheduling platform performs periodic spatiotemporal data mining through a heatmap generation module for sudden events. Massive amounts of historical data within a preset period (e.g., the past 3 months) are cleaned to extract the geographical coordinates (latitude and longitude) and timestamps of each overflow alarm event. Subsequently, instead of using a simple grid counting method, a spatiotemporal kernel density estimation algorithm is applied. This algorithm uses a Gaussian kernel function to smooth discrete alarm points in both spatial and temporal dimensions, generating a continuous probability density surface. Through this processing, high-frequency hotspot areas with significantly higher overflow probabilities than surrounding areas during specific time periods (such as Friday evening rush hour or the summer fruit season) can be identified, and the dynamic boundary contours of these hotspot areas can be accurately delineated.

[0049] Based on the aforementioned heatmap, when the current time enters the pre-defined lead time of the active period in the hotspot area (e.g., 15 minutes before the active period), the cloud-based scheduling platform retrieves all personnel located within the buffer zone at the edge of that hotspot area, filters out candidates currently in an idle or low-load return-to-work state, and issues a cruise guidance command at this time. This command does not specify a particular trash can, but instead calculates the center of gravity based on the probability density distribution within the hotspot area and generates a cruise route pointing to that center of gravity on the map of the personnel's smart wearable device.

[0050] To achieve the shortest physical response distance, the high-probability center of the hotspot area is set as a virtual gravitational source, and the workers are set as the force-bearing point masses. By calculating the virtual gravitational field, workers are guided to perform centripetal maneuvers from the edge of the area towards the center, ensuring that they are already at the physical geometric center or high-risk section of the area before an actual overflow incident occurs. Once any trash can in the heat map prediction area actually triggers an overflow alarm, the physical distance between the workers and the target point has been minimized (usually only a few hundred meters), thereby drastically reducing the average response time for emergencies from minutes to seconds, achieving a tactical upgrade from reactive firefighting to proactive control.

[0051] In some embodiments, the smart trash can is also equipped with an odor gas sensor, which is also used to upload odor concentration. The cloud-based scheduling platform is also used to determine the urgency of the cleanup. , where α and β are weighting coefficients, the regional background reference value is uploaded in real time by environmental monitoring nodes deployed in the same area, and is used to filter out environmental background odor noise; when the cleaning urgency result S exceeds the preset alarm baseline, even if the filling rate has not reached the physical overflow state, the overflow alarm signal is forcibly triggered and marked as a high-priority deodorization task in the generated composite task scheduling instruction.

[0052] In this embodiment, the smart trash can serves not only as a capacity monitoring terminal but also as an olfactory sensing node. It integrates an odor gas sensor specifically designed to monitor characteristic gases produced by garbage decay (such as ammonia, hydrogen sulfide, and volatile organic compounds). To address the issue of fluctuations in single concentration values, the raw voltage signal uploaded by the odor gas sensor is first subjected to baseline drift calibration and sliding window filtering to eliminate transient interference caused by pedestrians smoking or vehicle exhaust, thus extracting a steady-state odor concentration value C representing the true degree of decay inside the trash can. odor Subsequently, the cloud-based scheduling platform determined the urgency of the cleanup. The weighting coefficients α and β are not static constants, and the regional background reference value is uploaded in real time to the cloud scheduling platform by environmental monitoring nodes deployed in the same area (such as air quality monitoring stations integrated with smart streetlights or independent environmental sensing nodes). Through this algorithm, the system can dynamically remove the background odor noise (such as emissions from surrounding chemical plants, dense vehicle exhaust, etc.) from the raw signals collected by the odor sensors, thereby accurately pinpointing the local odor surge caused by garbage decay.

[0053] Weighting coefficients and It is not a static constant, but rather depends on ambient temperature and relative humidity. H The dynamic variable of the hook. When the ambient temperature rises (e.g.) The rate of odor diffusion increases exponentially, automatically increasing the odor weighting coefficient $\beta$ (e.g., from 0.3 to 0.6), while simultaneously decreasing the weighting of the fill rate. In the hot summer, even if the trash can is only half full, the urgency of cleaning it increases as soon as it produces an odor. It can also surge rapidly; while in the cold winter, it reverts to a scheduling logic centered on the fill rate.

[0054] Based on the calculated urgency of the cleanup A virtual overflow triggering mechanism was introduced, and an alarm baseline S was set. threshold (For example, 80 points). When S>S threshold Even if the physical filling monitoring sensor shows that there is still 50% space left in the bin, it will still be forcibly judged as being in a logical overflow state, immediately triggering the highest level overflow alarm signal. In the generated composite task scheduling instruction, this node will be marked with a special biochemical hazard label. After receiving this instruction, the smart wearable device will not only plan the optimal route, but also pop up a high-priority deodorization task prompt, requiring the operators to perform additional disinfection or deodorizing agent spraying procedures after cleaning, and take and upload photos of the inside of the bin after deodorization as a closed-loop acceptance certificate.

[0055] In some embodiments, the smart wearable device is a smart helmet or AR glasses; The smart wearable device is also used to overlay and project the emergency cleaning route and the location of the associated node onto the real scene on the lens display of the smart helmet or AR glasses; when the operator approaches the target trash can or associated node, the current filling percentage and the type of trash inside the trash can are displayed intuitively on the display in the form of augmented reality (AR) tags, to help the operator prepare the corresponding cleaning tools in advance.

[0056] In this embodiment, the smart wearable device worn by the worker integrates industrial-grade AR glasses or a smart helmet with an optical waveguide (display module and inertial measurement unit (IMU)). Combining GPS positioning data with the visual odometry of the front-facing camera, the 6-DoF pose of the worker's head in three-dimensional space is calculated in real time. The emergency cleanup route issued by the cloud dispatch platform is no longer a planar line on a two-dimensional map, but is transformed into a virtual guide light strip in three-dimensional space by the rendering engine. Through coordinate system registration technology, the virtual light strip is accurately anchored to the road surface in the real world. No matter how the worker turns their head, the navigation guidance seen through the AR lenses always closely follows the actual road, realizing a WYSIWYG immersive navigation that completely frees the worker's hands, eliminating the need for them to be distracted by looking at the handheld terminal.

[0057] When operators move to within a preset line-of-sight range (e.g., 30 meters) of the target trash can or associated node, the front-facing camera captures an image of the trash can in the real-world scene. Based on the correspondence between the pixel coordinates of the smart trash can's 2D image and its 3D world coordinates, the precise depth and angle of the smart trash can relative to the AR glasses are calculated. Subsequently, the rendering engine generates a floating AR augmented reality label at the corresponding position on the AR display screen. This label uses visual occlusion processing, making it appear as if it is floating above the physical trash can. This AR label not only displays a digital fill percentage, but more importantly, it intuitively shows the type of waste inside through color coding and icon mapping. For example, when the label detects that the trash can contains kitchen waste and the fill rate is too high, it turns green and includes a high-density / heavy object icon, prompting operators to prepare leak-proof hooks and a pushcart in advance. When hazardous waste or excessive odor is detected, the label immediately switches to a bright red warning and displays a protective glove / mask icon. This allows operators to complete their mental preparation and tool selection decisions during the last few meters of approach before physically touching the trash can, thus significantly improving the smoothness and safety of on-site operations.

[0058] In some embodiments, after sending the composite task scheduling instruction, the cloud scheduling platform continuously receives the location coordinates of the smart wearable device; when it detects that the real-time movement trajectory of the target worker continuously deviates from the emergency cleanup route by more than the allowable threshold, or that the dwell time at an associated node is abnormal, it sends an alarm to the administrator, sends a reminder inquiry to the smart wearable device, or initiates a backup worker matching process.

[0059] In this embodiment, after issuing the composite task scheduling command, the cloud scheduling platform does not disconnect. Instead, it continuously receives GPS / BeiDou dual-mode positioning coordinates uploaded by the smart wearable device at a high frequency (e.g., 1Hz) via WebSocket or MQTT long connection protocol. To determine deviations, the emergency cleanup route is abstracted as a series of connected vector line segments, and the current coordinates P of the worker are calculated in real time. cur To the nearest planned path segment L route Vertical projection distance D xte Set a fault tolerance buffer threshold to filter out false alarms caused by GPS signal drift or road fine-tuning. Only when D xte Only when the deviation consistently exceeds this threshold and lasts for more than a preset deviation confirmation window (e.g., 3 consecutive minutes) is it determined that the operator has committed a substantial deviation or dereliction of duty. At this point, instead of immediately notifying the administrator, a first-level intervention is triggered: a vibration and voice prompt (e.g., "You have deviated from the planned route, please correct immediately") is sent to the smart wearable device, and the correct return arrow is displayed on the AR glasses or screen. Simultaneously, regarding the determination of abnormal dwell time at associated node locations, a timer is automatically started when the operator enters the geofence (e.g., a radius of 15 meters) of a specific smart trash can. The average time T of the past N cleaning operations for that specific trash can is retrieved. avg and standard deviation σ. Set an outlier threshold T. limit =T avg +3σ. If the worker's stay at this node exceeds T... limit (For example, if it usually takes 5 minutes to clean up, but now it has been 20 minutes), it is determined that there may have been an accidental injury, vehicle malfunction, or an unauthorized extended rest period.

[0060] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention.

Claims

1. A smart sanitation emergency cleaning system based on multi-terminal collaboration, characterized in that, The system includes: a number of smart trash cans distributed in urban areas, a number of smart wearable devices worn by operators, and a cloud-based dispatch platform that communicates with the smart trash cans and the smart wearable devices; The smart trash can is equipped with a filling monitoring sensor and a first communication module, which is used to monitor the filling status data inside the can in real time, and send an overflow alarm signal to the cloud scheduling platform when the filling status data reaches the overflow threshold. The smart wearable device is equipped with a positioning module and a second communication module, which are used to collect the location coordinates of the workers in real time and upload them to the cloud dispatch platform, as well as to receive and display instructions from the cloud dispatch platform in a tactile or visual manner. The cloud-based scheduling platform is used to respond to the received overflow alarm signal by locking the corresponding smart trash can as the target trash can; matching the target worker within the response range based on the location coordinates uploaded by the smart wearable device; generating an emergency cleaning route based on the location coordinates of the target worker and the target trash can; establishing a spatial buffer zone of a preset width with the emergency cleaning route as the axis, and retrieving other smart trash cans located within the spatial buffer zone as associated nodes; obtaining the current filling status data of the associated nodes, and adding the associated nodes to the cleaning queue if the preset route-following cleaning conditions are met; generating a composite task scheduling instruction containing the target trash can and the associated nodes in the cleaning queue, and sending it to the smart wearable device of the target worker.

2. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The cloud-based dispatch platform is specifically used to acquire traffic topology data of the roads traversed by the emergency clearing route, identify the direction of travel and median strip information; based on the traffic topology data, eliminate areas that are physically adjacent to the emergency clearing route but cannot be legally changed lanes to enter directly; determine a set of areas that can be reached and returned to the original route from any node on the emergency clearing route within a preset additional travel time threshold, and construct the set of areas as the spatial buffer zone; the cloud-based dispatch platform is also used to acquire dynamic road condition data related to the emergency clearing route in real time, and dynamically scale the preset width of the spatial buffer zone and the additional travel time threshold according to the road congestion index; wherein, the conditions for clearing along the route include the road lateral direction of the associated node being consistent with the current travel direction of the workers, or the existence of a U-turn intersection at a distance less than a preset value.

3. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The cloud-based dispatch platform is specifically used to, when matching the target workers within the response range, also receive the current load status data of the vehicles driven by the workers uploaded by the smart wearable device; analyze the historical average compression density contained in the overflow alarm signal of the target trash can to estimate the weight of the target trash; calculate the remaining load capacity of the vehicle, and only when the remaining load capacity is greater than the sum of the weight of the target trash and the preset safety redundancy, mark the corresponding workers as candidate target workers; if all workers within the response range do not meet the capacity verification steps, a relay cleaning instruction is generated, and an empty vehicle is dispatched to provide support.

4. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The smart wearable device is equipped with a near-field communication module; The smart trash can is equipped with a corresponding electronic tag recognition module; The smart wearable device is also used to perform near-field handshake communication with the electronic tag recognition module of the smart trash can through the near-field communication module when the target worker arrives at the target trash can or the associated node location for cleaning, and to generate a human-trash can interaction verification code. The smart trash can is also used to generate an emptying reset signal after it is determined to be emptied by monitoring sensors; The cloud-based scheduling platform is also used to determine that the cleaning task of the smart trash can has been completed when it simultaneously receives an emptying and reset signal from the smart trash can and a human-trash can interaction verification code generated from near-field handshake communication, and automatically removes the node corresponding to the smart trash can from the navigation queue on the smart wearable device.

5. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The filling monitoring sensor of the smart trash can includes a multispectral camera component and a visible light camera component; The cloud-based scheduling platform is also used to identify surface coverings in images captured by the visible light camera component before generating composite task scheduling instructions, and to determine whether there are any obstructions that hinder spectral penetration; if there are no obstructions, spectral analysis is performed on the garbage images in the target garbage bin and the associated node to identify the hazard level and recycling category of the garbage content; the loading type attribute of the vehicle equipped by the target operator is obtained; if the garbage recycling category of the associated node does not match the loading type attribute, even if the associated node is located within the spatial buffer zone, it is prohibited from being added to the cleaning queue, and a sorting and transfer task is generated.

6. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The smart trash can is equipped with an adaptive frequency adjustment module, which is also used to monitor the rate of change of filling status data in real time. When the rate of change exceeds a preset sudden growth threshold, the sampling frequency of the first communication module sending data to the cloud scheduling platform is increased. When the filling status data is detected to remain unchanged for a preset duration, it enters a low-power sleep mode and sends status data only at a preset heartbeat time each day. The smart trash can also includes a power monitoring module, and the cloud scheduling platform dynamically adjusts the upper limit of the sampling frequency increase based on the remaining power fed back by the power monitoring module. The cloud-based scheduling platform is also used to predict the remaining time window before the smart trash can reaches the overflow state based on the rate of change, and to issue the overflow alarm signal in advance before the remaining time window is less than the time required for the operator to arrive.

7. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The cloud-based dispatch platform also includes a heat map generation module for sudden events, which is used to statistically analyze the spatiotemporal distribution of overflow alarm signals issued by smart trash cans in various areas within a preset period; identify hotspot areas with high frequency of overflow; and send pre-dispatch instructions to the smart wearable devices of idle workers located at the edge of the hotspot area during the time period corresponding to the hotspot area, so as to guide the workers to move towards the center of the hotspot area for patrol, thereby shortening the physical distance to respond to sudden overflow events.

8. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The smart trash can is also equipped with an odor gas sensor, which is used to upload odor concentration data. The cloud-based scheduling platform is also used to determine the urgency of the cleanup. , where α and β are weighting coefficients, the regional background reference value is uploaded in real time by environmental monitoring nodes deployed in the same area, and is used to filter out environmental background odor noise; when the cleaning urgency result S exceeds the preset alarm baseline, even if the filling rate has not reached the physical overflow state, the overflow alarm signal is forcibly triggered, and it is marked as a high-priority deodorization task in the generated composite task scheduling instruction.

9. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, The smart wearable device is a smart helmet or AR glasses; The smart wearable device is also used to overlay and project the emergency cleaning route and the location of the associated nodes onto the real-world scene on the lens display of the smart helmet or AR glasses. The smart wearable device uses a fusion positioning algorithm, combined with visual odometry and inertial measurement unit, to compensate and correct the projection position of the emergency cleaning route using inertial navigation data when environmental visual features are lost or GNSS signals are weak. When the operator approaches the target trash can or associated node, the current filling percentage and internal trash type of the trash can are displayed intuitively on the display in the form of augmented reality (AR) tags, assisting the operator in preparing the corresponding cleaning tools in advance.

10. The intelligent sanitation emergency cleaning system based on multi-terminal collaboration according to claim 1, characterized in that, After sending the composite task scheduling instruction, the cloud scheduling platform continuously receives the location coordinates of the smart wearable device. When it detects that the real-time movement trajectory of the target worker deviates from the emergency cleanup route by more than the allowable threshold, or that the dwell time at an associated node is abnormal, it sends an alarm to the administrator, sends a reminder inquiry to the smart wearable device, or initiates the backup worker matching process.

Images