A system, method and computer readable storage medium for online detection and rejection of a storage and retrieval tray

Abstract

This application discloses an online detection and rejection system, method, and computer-readable storage medium for automated warehouse pallets. The system includes: a detection module for acquiring three-dimensional geometric data of pallets to be received online; a data processing and judgment module for comparing the acquired three-dimensional geometric data with preset standard pallet data, and determining an abnormal pallet and generating an abnormal signal when the detection deviation exceeds a preset threshold; this module can also evaluate structural integrity based on partition weighting and calculate the center of gravity offset vector; a scheduling system for receiving abnormal signals and locking abnormal pallets; and execution equipment, including automated guided vehicles or sorting mechanisms, for removing abnormal pallets according to the schedule. This application enables high-precision online detection and quantitative damage assessment of incoming pallets, and adaptive safety control of the handling process based on defect data such as center of gravity offset, improving detection accuracy and the overall safety of automated warehousing operations.

Description

Technical Field

[0001] This application relates to the field of automated warehousing technology, and more specifically, to a system, method, and computer-readable storage medium for online pallet status detection, intelligent assessment, and collaborative handling control in automated storage and retrieval systems (AS / RS). Background Technology

[0002] Automated storage and retrieval systems (AS / RS), as a core component of modern logistics systems, achieve high-density storage and automated retrieval of goods through high-rise racking, stacker cranes, inbound and outbound conveyor systems, and automated control systems. Pallets, as the most basic load-bearing unit, are essential for the stable and efficient operation of the entire AS / RS system, and their structural integrity and dimensional standardization are prerequisites for this.

[0003] In actual warehousing operations, pallets inevitably suffer from wear, cracks, and deformation during frequent turnover and handling. If a damaged or severely deformed pallet enters an automated storage and retrieval system (AS / RS) without inspection, it can lead to a series of serious consequences. For example, damaged pallet support pads may cause instability on the racks, resulting in goods tipping over or scattering, threatening operational safety. Dimensional deviations or deformation of the pallet may cause it to jam when stacker cranes are handling goods in aisles, even damaging the cranes or racks, leading to equipment malfunctions and prolonged downtime of the entire warehousing system, severely impacting production and logistics efficiency.

[0004] Existing solutions typically involve setting up simple inspection stations at the entrance of automated warehouses, such as using photoelectric sensors or mechanical barriers to detect excessive height and width. While these methods can intercept some severely oversized pallets, they have several shortcomings: First, the detection accuracy is limited, failing to identify subtle but critical defects such as localized damage, panel cracks, or structural damage to the pallet. Second, the judgment logic is simplistic, usually only performing a binary judgment of pass or fail, lacking a quantitative assessment of the severity and type of damage. Third, when handling non-compliant pallets, rigid mechanical rejection methods are usually used, resulting in a rigid process that does not consider the secondary safety risks that pallet defects may pose to subsequent rejection and handling processes, such as the pallet scattering or tipping over during transport.

[0005] Therefore, how to achieve online, accurate, and quantitative defect detection of incoming pallets, and how to safely and flexibly handle abnormal pallets based on the detection results, is a technical problem that urgently needs to be solved in the field of automated warehousing. Summary of the Invention

[0006] The purpose of this application is to address the problems in the existing technology, such as insufficient detection accuracy of incoming pallets, inability to quantify damage assessment, and secondary safety risks in the abnormal handling process, and to provide an online detection and rejection system, method, and computer-readable storage medium for automated warehouse pallets.

[0007] To achieve the above objectives, this application provides an online detection and rejection system for automated storage and retrieval systems (AS / RS) pallets, comprising: a detection module installed on the entrance conveyor line of the AS / RS for acquiring three-dimensional geometric data of pallets to be received online; a data processing and judgment module connected to the detection module for comparing the acquired three-dimensional geometric data with preset standard pallet data, and determining the pallet to be received as an abnormal pallet and generating an abnormal signal when the detection deviation exceeds a preset threshold; a scheduling system communicating with the data processing and judgment module for receiving the abnormal signal; and an execution device, including an automated guided vehicle (AGV) or a sorting mechanism; upon receiving the abnormal signal, the scheduling system locks the abnormal pallet and schedules the execution device to move the abnormal pallet from the entrance conveyor line to a preset abnormal handling area.

[0008] This application also provides an online detection and rejection method for automated warehouse pallets, applied to the aforementioned system, comprising the following steps: acquiring three-dimensional geometric data of pallets to be received on the inlet conveyor line; comparing the three-dimensional geometric data with preset standard pallet data, and determining the pallet to be received as an abnormal pallet and generating an abnormal signal when the detection deviation is greater than a preset threshold; receiving the abnormal signal through a scheduling system and locking the abnormal pallet; and a scheduling execution device moving the abnormal pallet from the inlet conveyor line to a preset abnormal processing area.

[0009] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method steps.

[0010] This application achieves a complete closed loop from high-precision online detection and intelligent damage assessment to flexible collaborative handling through an integrated system architecture. Compared with existing technologies, this application has the following advantages: By constructing a 3D point cloud model of the pallet and introducing a partitioned weighted structural integrity assessment algorithm, it can accurately identify structural damage that substantially affects load-bearing capacity, while filtering out false alarms caused by non-critical defects such as minor surface wear, increasing the identification rate of true structural damage to over 98%. This application uses the detected center of gravity offset vector data to guide the handling process of the execution equipment. By dynamically adjusting the kinematic parameters of the automated guided vehicle, it achieves safe and adaptive handling of damaged pallets, actively avoiding secondary stacking or tipping accidents caused by handling pallets with unstable centers of gravity at standard speeds. In addition, the system is deployed on the main conveyor line, enabling continuous detection without stopping, with a processing capacity of over 120 pallets / hour. Simultaneously, the upper-level scheduling system flexibly rejects abnormal pallets, avoiding prolonged shutdowns of the main warehousing process due to abnormal handling, and ensuring the operational efficiency of the entire warehousing system. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic diagram of the architecture of an online detection and rejection system for automated warehouse pallets provided in an embodiment of this application.

[0013] Figure 2 This is a schematic diagram of the process flow for an online detection and rejection method for automated warehouse pallets provided in an embodiment of this application.

[0014] Figure 3 This is an internal logic block diagram of the data processing and judgment module provided in the embodiments of this application.

[0015] In the diagram: 100 - inlet conveyor line, 101 - detection module, 102 - data processing and judgment module, 103 - scheduling system, 104 - execution equipment. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0017] Example 1

[0018] This embodiment provides an online detection and rejection system and method for automated storage and retrieval systems (AS / RS) pallets. It is designed to achieve efficient and accurate quality screening of pallets entering AS / RS and to safely and intelligently reject unqualified pallets.

[0019] Reference Figure 1 In this embodiment, the online pallet detection and rejection system for automated storage and retrieval systems (AS / RS) is deployed in the conveyor line section before the entrance of the AS / RS. The system physically includes an entrance conveyor line 100, a detection module 101, a data processing and judgment module 102, a scheduling system 103, and an execution device 104.

[0020] The inlet conveyor line 100 is preferably a multi-segment continuously distributed powered roller conveyor or chain conveyor, responsible for transporting pallets with stacked goods from the upstream workstation to the inspection area and the subsequent automated warehouse entrance. Positioning components, such as mechanical stops or photoelectric sensors, can be installed on the conveyor line segment where the inspection module 101 is located to ensure that the pallet is in a preset reference position when entering the inspection area, thereby improving inspection accuracy.

[0021] The detection module 101 is used to acquire the three-dimensional geometric data of the pallet to be received online. In this embodiment, the detection module 101 is specifically embodied as a gantry-type detection rack that spans above the inlet conveyor line 100. A composite sensor is integrated on the rack. Specifically, the composite sensor may include a machine vision component, a laser scanning component, and a pulse displacement sensor. A preferred configuration is to install three sets of 3D line laser profilometers on the top and sides of the rack. The scanning field of view of the three sets of 3D line laser profilometers can completely cover the top and four edges of the pallet on the conveyor. In order to effectively acquire the load-bearing structure information at the bottom of the pallet, the optical axis of the 3D line laser profilometers on both sides is tilted downwards at an angle of 15 to 20 degrees to the horizontal plane to ensure that the load-bearing pads and forklift insertion areas at the bottom of the pallet can be scanned. In addition, the detection module 101 may also integrate a barcode reader to acquire the unique identification mark affixed to the pallet during detection and bind the pallet ID with the detection result.

[0022] The data processing and judgment module 102 is typically a local edge computing controller connected to the sensors of the detection module 101 via an industrial Ethernet network. Its core function is to process raw sensor data and determine defects. Its workflow is as follows: when a pallet to be received moves along the inlet conveyor line 100, a photoelectric trigger switch installed at the entrance of the detection rack is activated, starting a 3D line laser profilometer for continuous profile scanning. The data processing and judgment module 102 receives and stitches these profile data, constructing a complete 3D point cloud model containing the pallet and its cargo in real time. After obtaining the original 3D point cloud model, the data processing and judgment module 102 first performs preprocessing. A Random Sample Consensus (RANSAC) plane fitting algorithm can be used to identify and filter out point clouds belonging to the conveyor line background, extracting a clean pallet point cloud. Next, to improve the accuracy of subsequent comparisons, noise reduction processing can be performed on the pallet point cloud, using a bilateral filtering algorithm to remove discrete noise points generated by factors such as reflection at the pallet edges. After preprocessing, the data processing and judgment module 102 registers the processed 3D point cloud model with a standard pallet computer-aided design (CAD) model pre-stored in the database. The registration algorithm can employ the Iterative Closest Point (ICP) algorithm. Through registration, the actual 3D geometric data of the pallet to be inspected, such as its external dimensions, surface flatness, and column tilt angle, can be accurately calculated, along with the deviation from the standard data.

[0023] The scheduling system 103 is the intelligent hub of the entire system, responsible for decision-making and coordination. It typically consists of a warehouse management system (WMS) and a robot scheduling system. The data processing and judgment module 102 establishes a communication connection with the warehouse management system via a network protocol. The warehouse management system, as the upper-level system, manages the inventory and operational tasks of the entire warehouse, while the robot scheduling system is mounted at its lower level. The robot scheduling system is responsible for the path planning and task scheduling of the specific execution equipment, and interacts with the controllers of the execution equipment 104 (such as automated guided vehicles (AGVs)) in the workshop via a wireless local area network.

[0024] The execution device 104 is the physical unit that performs the rejection task. In this embodiment, the execution device 104 is preferably an automated guided vehicle (AGV), which is equipped with a lifting mechanism, such as a hydraulic or electric lifting platform, capable of lifting and transporting the abnormal pallet as a whole. Downstream of the inspection station of the inlet conveyor line 100, a diversion node is set up, which connects the main inlet line to a preset abnormality handling area. The abnormality handling area can be further divided into a maintenance workstation and a waste recycling station for manual handling.

[0025] Reference Figure 2 and Figure 3 The complete working method of this embodiment is as follows: When the pallet enters the detection area, the photoelectric switch triggers the detection module 101, the 3D line laser profilometer scans the pallet, and sends the raw signal to the data processing and judgment module 102, which constructs a three-dimensional point cloud model of the pallet. Afterwards, data comparison and defect judgment are performed. This judgment not only performs a simple dimensional deviation comparison but also introduces a structural integrity evaluation logic based on partition weighting. In this embodiment, a three-dimensional rectangular coordinate system is pre-established: the horizontal conveying surface of the inlet conveyor line 100 is the XY plane, and the direction perpendicular to the horizontal conveying surface and upward is the Z-axis direction. Specifically, the data processing and judgment module 102, based on the Z-axis height threshold, divides the pallet point cloud into a set of bottom load-bearing areas. and top panel area collection The load-bearing pad area corresponds to the nine independent block structures in the pallet that directly support the load, while the panel area corresponds to the horizontally laid panels connecting the tops of these load-bearing pads. Subsequently, the data processing and judgment module 102 calculates the volume loss in each load-bearing pad area. and the volume loss of the panel area The structural integrity of a pallet is quantified by assigning higher weights to load-bearing pad areas that have a greater impact on pallet load-bearing safety. This assessment logic uses structural integrity indicators... To illustrate this, the calculation logic is as follows: ,in, As the preset first weight, The second weight is preset, and In a specific application scenario, it can be set =0.8, =0.2. When the calculated structural integrity index... When the damage threshold is exceeded, the data processing and judgment module 102 determines that the pallet is an abnormal pallet and generates an abnormal signal. Through this weighted evaluation, the system can effectively distinguish between severe damage to load-bearing parts and minor wear on the panel, avoiding false alarms caused by non-structural defects such as wrapping film burrs or panel scratches.

[0026] If the pallet is determined to be abnormal, the data processing and determination module 102 will also perform two key actions. First, it calculates the actual center of gravity of the abnormal pallet based on the 3D point cloud model and compares it with the geometric center of a standard pallet to obtain a quantified center of gravity offset vector. This vector This characterizes the degree and direction of the pallet's center of gravity deviation. Secondly, it extracts feature images of abnormal pallets, such as magnified views of the damaged areas, and stores them in a database for traceability. Finally, the data processing and judgment module 102 will include the status code, pallet ID, and center of gravity offset vector. The data frames are sent to the scheduling system 103 via the network. The warehouse management system (i.e., the attached system) in the scheduling system 103... Figure 1 The warehouse management system (WMS) receives a data frame. If the status code is normal, the inlet conveyor line 100 is instructed to continue operating and release the pallet to the entrance of the automated warehouse. If the status code is abnormal, the warehouse management system immediately locks the logistics status of the abnormal pallet in the system, preventing it from entering the subsequent standard inbound process, and issues an audible and visual alarm signal via an alarm light.

[0027] Subsequently, the warehouse management system issues the rejection task to the robot scheduling system. Based on the current position of the defective pallet and the location of the preset defect handling area, the robot scheduling system plans an optimal transport path for an idle automated guided vehicle (AGV) and sends instructions containing path planning and kinematic parameters to the execution device 104, i.e., the AGV. This process achieves kinematic adaptive scheduling controlled by defect data. In conventional scheduling systems, AGVs execute all tasks using fixed preset acceleration and deceleration parameters. In this scheme, the scheduling system 103 receives the center of gravity offset vector. Then, it will be based on its modulus. The motion control parameters of the automated guided vehicle are dynamically derated. Specifically, the motion control parameters include the maximum acceleration limit. and maximum centripetal force limit The kinematic engine within the robot scheduling system will determine the magnitude based on the kinematic engine. The size is adjusted in real time to change the maximum acceleration limit for the current transport task. The adjustment logic satisfies: when the modulus... When increased, the maximum acceleration limit value The corresponding reduction. In a preferred embodiment, the adjustment logic may follow the following formula: ,in, It is the acceleration due to gravity. Let be the coefficient of friction between the contact surface between the pallet and the automated guided vehicle. For the width of the pallet, This represents the projection magnitude of the centroid offset vector in the horizontal direction. This is a safety redundancy constant. Furthermore, the scheduling system 103 can also adjust the modulus... A smoothing factor is calculated for the speed planning curve, which is used to limit the acceleration of the automated guided vehicle during start-up and stopping phases. For the steering process, the robot scheduling system bases its calculations on the center of gravity offset vector. The projection direction in the Automated Guided Vehicle (AGV) coordinate system limits the maximum angular velocity of the AGV during the turning process, thereby reducing the required centripetal force.

[0028] By coupling the three-dimensional geometric features of the detection end with the kinematic parameters of the execution end, this design enables adaptive and safe handling of damaged pallets. It effectively avoids the risk of secondary slippage or tipping when handling pallets whose center of gravity has already shifted significantly at default speeds. Finally, following instructions from the robot scheduling system, the automated guided vehicle (AGV) travels to the designated docking point, lifts and removes the abnormal pallet, and transports it to the anomaly handling area, awaiting human intervention. After completing the task, the AAV reports its status back to the robot scheduling system, becoming a usable resource again.

[0029] Example 2

[0030] This embodiment is a variant based on Embodiment 1. In some scenarios where processing speed requirements are extremely high but flexibility requirements are relatively low, the execution device 104 may adopt different configurations.

[0031] In this embodiment, the execution device 104 is not an automated guided vehicle (AGV), but a fixed sorting mechanism set at the diversion node. This sorting mechanism can be a fast-response pneumatic pusher device or a conveyor belt that can be raised, lowered, and moved laterally.

[0032] When the scheduling system 103 receives an abnormal signal, it stops scheduling the automated guided vehicle and instead sends a rejection command directly to the controller (such as a programmable logic controller) of the sorting mechanism. Upon receiving the command, the sorting mechanism quickly moves the abnormal pallet from the main conveyor line to the adjacent abnormal buffer conveyor line when the abnormal pallet moves into its range.

[0033] The advantages of this method are its extremely rapid rejection process, its ability to match very high mainline conveyor speeds, and its simple structure and low cost. The disadvantage is its lack of flexibility; defective pallets can only be transferred to a fixed buffer line, requiring further transfer by other equipment. This configuration is suitable for situations where the defective pallet occurrence rate is low and only rapid separation of defective items is needed, without the need for complex path planning. Although the execution equipment differs, the front-end detection module, data processing and judgment module, and their core weighted evaluation algorithm remain consistent with Example 1.

[0034] In summary, this application provides an integrated and intelligent online inspection and rejection solution for automated storage and retrieval systems (AS / RS). Through high-precision 3D visual inspection, physically-based zone-weighted damage assessment, and deep coupling of inspection results with the kinematic parameters of the handling equipment, it not only improves the accuracy and reliability of defect detection, but more importantly, it elevates the safety concept from passive interception to proactive prevention, providing strong technical support for the safe, stable, and efficient operation of automated warehousing systems.

[0035] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An online detection and rejection system for automated storage and retrieval systems (AS / RS), characterized in that, include: The detection module is installed on the entrance conveyor line of the automated warehouse to acquire the three-dimensional geometric data of the pallets to be put into storage online. The data processing and judgment module, connected to the detection module, is configured to construct a three-dimensional point cloud model of the pallet to be received based on the three-dimensional geometric data; divide the three-dimensional point cloud model into a load-bearing pad area and a panel area, and calculate a structural integrity index based on the volume loss of the load-bearing pad area and the panel area; when the structural integrity index exceeds a preset damage threshold, it is judged as an abnormal pallet and an abnormal signal is generated; the data processing and judgment module is also used to calculate the center of gravity offset vector of the abnormal pallet relative to the standard geometric center; The scheduling system is used to receive the abnormal signal and the center of gravity offset vector, and to lock the abnormal tray; The execution equipment includes an automated guided vehicle for removing the abnormal pallet according to instructions from the scheduling system; The scheduling system dynamically adjusts the motion control parameters of the automated guided vehicle when transporting the abnormal pallet based on the magnitude of the center of gravity offset vector.

2. The online detection and rejection system for automated warehouse pallets according to claim 1, characterized in that, The detection module is installed on the gantry structure above the inlet conveyor line and includes a machine vision component, a laser scanning component, a pulse displacement sensor, and a barcode reader.

3. The online detection and rejection system for automated warehouse pallets according to claim 1, characterized in that, The formula for calculating the structural integrity index by the data processing and judgment module is: , where is the volume missing amount of the load-bearing pad block area, is the volume missing amount of the panel area, is the preset first weight, is the preset second weight, and the load-bearing pad block area is given a higher weight by setting.

4. The online detection and rejection system for automated warehouse pallets according to claim 1, characterized in that, The data processing and judgment module is also used to remove discrete noise points from the three-dimensional point cloud model using a bilateral filtering algorithm, and to register the denoised three-dimensional point cloud model with a preset standard tray three-dimensional model using an iterative nearest-point algorithm.

5. The online detection and rejection system for automated warehouse pallets according to claim 1, characterized in that, The motion control parameters include a maximum acceleration limit value. The dynamic adjustment logic of the scheduling system is as follows: when the modulus increases, the maximum acceleration limit value decreases accordingly. The scheduling system also calculates the smoothing factor of the speed planning curve of the automated guided vehicle based on the modulus value.

6. The online detection and rejection system for automated warehouse pallets according to claim 5, characterized in that, The motion control parameters also include a maximum centripetal force limit value; the scheduling system also limits the angular velocity of the automated guided vehicle during the turning process based on the projection direction of the center of gravity offset vector in the coordinate system of the automated guided vehicle.

7. The online detection and rejection system for automated warehouse pallets according to claim 1, characterized in that, The scheduling system includes a warehouse management system and a robot scheduling system; the automated guided vehicle is equipped with a lifting mechanism for lifting and transporting the abnormal pallet; the scheduling system also issues an audible and visual alarm signal after the abnormal pallet enters the preset abnormal handling area.

8. The online detection and rejection system for automated warehouse pallets according to claim 1, characterized in that, The data processing and judgment module is also used to extract a feature image containing a magnified view of the damaged location of the pallet to be put into storage and save it to the database when the abnormal signal is generated.

9. A method for online detection and rejection of pallets in automated storage and retrieval systems, applied to the system described in any one of claims 1-8, characterized in that, include: Acquire the 3D geometric data of the pallets to be put into storage deployed on the entrance conveyor line and construct a 3D point cloud model; The three-dimensional point cloud model is divided into a load-bearing pad area and a panel area. The structural integrity is determined based on the weighted volume missing amount, and an abnormal signal is generated when the pallet is determined to be abnormal. Calculate the center-of-gravity offset vector of the abnormal pallet; The system receives the abnormal signal through the scheduling system, locks the abnormal pallet, and dynamically adjusts the motion control parameters of the automated guided vehicle according to the magnitude of the center of gravity offset vector. The automated guided vehicle is dispatched to move the abnormal pallet away according to the adjusted motion control parameters.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method of claim 9.

Images