Process supervision visualization method for e-commerce warehouse sorting operation

By building visual tools for picking jobs, including job performance overview, timing analysis and order process comparison view, the problem of insufficient multi-level and time-dimensional analysis in e-commerce warehouse picking job supervision is solved, and in-depth insights into picking jobs and Optimization support improves the efficiency and process control of picking operations.

CN119090395APending Publication Date: 2024-12-06SICHUAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411066114.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

It is difficult for the existing technology to effectively supervise the e-commerce warehouse picking process, especially in the multi-level analysis and exploration of dynamic changes in the time dimension, resulting in a lack of macro-understanding and detailed insights into the picking operation.

Method used

By obtaining and cleaning the picking process data, we build job performance overview view, job timing analysis view and order process comparison view. Combined with interactive response methods, we provide visual tools to assist administrators in supervising picking jobs from overall to detail, revealing the time-changing of picking jobs. Rules and abnormal phenomena.

Benefits of technology

It realizes comprehensive supervision and optimization decision support for picking operations, and improves the understanding and control of picking operations efficiency and process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119090395A_ABST
    Figure CN119090395A_ABST
Patent Text Reader

Abstract

The invention provides a process supervision visualization method for e-commerce warehouse sorting operation, and relates to the technical field of information visualization and visual analysis. The method comprises the following steps: acquiring manual sorting process data, and storing the cleaned and converted manual sorting process data into a database; the method comprises the following steps of: visually mapping manual sorting process data through a visual channel, namely, constructing an operation performance overview view, an operation time sequence analysis view and an order flow comparison view; dividing the canvas into a plurality of areas, and respectively arranging a job performance overview view, a job time sequence analysis view and an order flow comparison view; and performing visual construction on the operation performance overview view, the operation time sequence analysis view and the order flow comparison view. The method can assist a warehouse administrator in comprehensively sensing the operation state of the picking operation, analyzing the time-varying mode of the picking operation, revealing the evolution process of the order state in the picking batch, discovering and analyzing the abnormal order processing process, and providing support for the warehouse administrator to make a production decision.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of information visualization and visual analysis, and in particular to a process supervision visualization method for picking operations in e-commerce warehouses. Background Art

[0002] At present, many small and medium-sized e-commerce companies still rely on traditional manual picking operations, which involve creating picking tasks, allocating inventory and generating picking batches, workers receiving picking tasks and picking goods, sorting goods and closing orders, etc., which is highly complex. Therefore, how to effectively supervise the picking process of e-commerce warehouses, discover potential efficiency problems, and propose optimization measures is an important problem in e-commerce warehouse management.

[0003] Currently, warehouse picking operation analysis mainly relies on models and algorithms to optimize key issues such as order batching. However, the above methods often ignore the rich experience and logical analysis capabilities of experts in the field, making it difficult for experts to understand the inherent logic of the model operation and lack a full understanding and trust in the optimization results. In response to the above shortcomings, visualization and visual analysis use concise and intuitive visualization charts to give users the ability to study data in depth in an interactive way. Although some visual analysis technologies can be used to explore patterns and rules in complex production processes, they mainly focus on analyzing scenarios with stable production workload and work progress, and have limited applicability in warehouse scenarios with high uncertainty. Specifically, there are the following shortcomings:

[0004] 1) In the picking operation, the management objects cover multiple levels such as picking batches, orders and goods. Existing research lacks multi-level analysis and exploration of picking operations. Some studies focus on overview analysis and ignore in-depth exploration of the details of picking operations, while some studies focus on detailed analysis and lack a macro understanding and grasp of the entire operation.

[0005] 2) The data of the picking operation process integrates complex time attributes and event attributes, including key events such as creation, allocation, picking, and closing. Existing visualization methods are difficult to effectively capture the dynamic changes of event dimensions in the time dimension, and cannot deeply reveal potential abnormal phenomena in picking operations. Summary of the invention

[0006] In view of the above problems, the purpose of the present invention is to provide a process supervision visualization method for picking operations in e-commerce warehouses, which can assist warehouse managers in realizing process supervision of picking operations from overall overview to detailed inspection, and complete the tasks of picking operation performance overview, picking operation time-varying law analysis, and order status evolution process analysis. Through this method, warehouse managers can deeply explore the implicit laws and patterns of picking operations, and provide decision support for optimizing picking operation processes and improving picking operation efficiency.

[0007] The specific technical solutions are as follows:

[0008] Step S1: Acquire manual picking process data, and store the cleaned and converted manual picking process data into a database;

[0009] Step S2: Visually map the manual picking process data through the visual channel, including: constructing an operation performance overview view, an operation timing analysis view, and an order process comparison view;

[0010] Step S3: Divide the canvas into multiple areas, and arrange the job performance overview view, job timing analysis view, and order process comparison view respectively;

[0011] Step S4: Based on the job performance overview view, the job timing analysis view and the order process comparison view, a visual interactive response method for viewing the manual picking process data is constructed.

[0012] Preferably, step S1 specifically includes:

[0013] Step S11: Obtaining a delivery order record table from the warehouse management system of the warehousing service provider;

[0014] The attributes of the outbound order record table include: picking batch code, order code, product code, picking personnel, creation timestamp, allocation timestamp, picking timestamp and order closing timestamp;

[0015] Step S12: data cleaning and data conversion;

[0016] Step S121: deleting erroneous data and missing data;

[0017] Step S122: converting the time attribute in the outbound order data from seconds to “year-month-day”;

[0018] Step S123: Counting four performance indicators of the warehouse's daily picking operation, including: the total number of picking batches, the total number of orders, the total number of picked goods, and the total number of picked goods types, to obtain operation performance overview data;

[0019] Step S124: Count the total number of picking batches in the four events of creation, allocation, picking and order closing that leave the warehouse every hour every day to obtain operation timing analysis data;

[0020] Step S125: Taking the picking batch as the analysis unit, setting multiple time slices, calculating the overall processing time of each picking batch, the time slice division unit, the number of time slice divisions, and the order status list corresponding to each time slice, and obtaining the order processing process data;

[0021] Step S13: Store the operation performance overview data, operation timing analysis data and order processing process data in the database.

[0022] Preferably, the database is a MySQL database.

[0023] Preferably,

[0024] Build a job performance overview view based on job performance overview data;

[0025] For each performance indicator, four colors are used to represent the minimum value, lower quartile, upper quartile and maximum value of a single day in the circular calendar chart; the values ​​of the corresponding performance indicators are linearly mapped between the four colors through a gradient color system, and each dot representing a single day is filled with a different color;

[0026] For each performance indicator, the nested band is divided into multiple sub-bands according to the number of months, and the outer boundary line, inner boundary line and middle line of the sub-band represent the maximum value, minimum value and median of the corresponding month respectively; in the nested radar chart, the radial scale axis represents the average value of the corresponding month.

[0027] Build a job timing analysis view based on job timing analysis data;

[0028] Rectangles of four colors represent creation, allocation, picking, and order closing in a single hour, and different lengths represent different quantities. The rectangles are stacked into a job timing analysis subgraph in the order of creation, allocation, picking, and order closing in units of hours, and arranged in chronological order.

[0029] Connect rectangles of the same color in each job timing analysis subgraph in time sequence in the form of strips.

[0030] Build order flow comparison view based on order processing data;

[0031] Four plane graphics are used to represent creation, allocation, picking and order closing, and the size of the plane graphics represents different quantities;

[0032] On the time coordinate axis, events in the same time slice are connected by straight lines and arranged vertically in order; adjacent plane figures on adjacent time slices are connected by cubic Bezier curves.

[0033] Preferably,

[0034] The placement job performance overview views include:

[0035] Calculate the fill color of the dots,

[0036]

[0037] In the formula, Hex iis the hexadecimal code of the color corresponding to the i-th performance indicator value; Value i is the i-th performance indicator value, Value min 、Value max They are the minimum and maximum values ​​of the performance indicators, Hex max 、Hex min They are the hexadecimal codes corresponding to the maximum and minimum values ​​of the performance indicators respectively;

[0038] Based on the predefined canvas boundaries, calculate the absolute coordinates of each point in the circular calendar chart.

[0039]

[0040]

[0041]

[0042] In the formula, θ radians is the angle that each month needs to be rotated in the canvas; j represents the month, j=1,2,...,N, N∈[1,2,3,...,12]; (x local ,y local ) is the relative coordinate of the dot; col represents the date number of the dot corresponding to the day of the week, col = 0, 1, 2, ... 6; row represents the week number of the dot corresponding to the day of the month, row = 0, 1, 2, ...; Radius is the radius of the dot, circleGap is the distance between the dots, verticalPos is the vertical distance between the dot and the top of the canvas, (x global ,y glocal ) is the absolute coordinate of the dot; s is the canvas scaling factor, 0 <s<1;

[0043] Draw dots according to their absolute coordinates and fill them with colors;

[0044] In the center of the annular calendar chart composed of multiple months, nested ring belts and nested radar charts are arranged from the outside to the inside;

[0045] Among them, multiple sub-bands are connected end to end to form a closed interval, and the dotted background of the band represents the radial scale; the nested radar chart includes three dotted concentric circles. From the inside to the outside, the first concentric circle represents the minimum value of the average value of multiple months, and the third concentric circle represents the maximum value of the average value of multiple months.

[0046] The job timing analysis view includes:

[0047] Calculate the starting point for drawing each job timing analysis subgraph,

[0048]

[0049]

[0050] In the formula, (rect x ,rect y ) is the coordinate of the drawing starting point, StackSvgWidth and StackSvgHeight are the width and height of the canvas respectively, bandGap is the interval of the band, k is the hour number, stackCnt is the number of stacked batches, initCnt is the number of original batches, maxBathSum is the maximum batch capacity of the canvas, and rectGap is the interval of the job timing analysis subgraph;

[0051] Draw each job timing analysis subgraph based on the drawing starting point.

[0052] The layout order process comparison view includes:

[0053] Determine the plane figure corresponding to the event,

[0054] stateSymbol processType =symbolGenerator(processType);

[0055] In the formula, processType is the event type, stateSymbol processType is the plane figure corresponding to the processType event, and symbolGenerator(·) is the generating function of the plane figure;

[0056] The domain of the linear scale is determined by the number of time slices corresponding to the picking batch processing process, and the range of the linear scale is determined by the canvas width; the X-axis coordinate of the time coordinate axis is determined by the linear scale;

[0057] Determine the Y-axis coordinate symbol of the time axis y ,

[0058] symbol y =startY+(l-1)×gapSpace;

[0059] Where startY is the starting position of the first event symbol in the lth time slice, and gapSpace is the spatial interval between the plane figures;

[0060] Calculate control point controlP1 and control point controlP2,

[0061]

[0062]

[0063] Where startPX l is the X-axis coordinate of the starting point of the lth time slice, startPY l ——The Y-axis coordinate of the starting point of the lth time slice, endPX l is the X-axis coordinate of the end point of the lth time slice, endPY l is the Y-axis coordinate of the end point of the lth time slice, and controlVal is the offset distance of the control point in the horizontal position;

[0064] Draw a cubic Bezier curve based on the starting point startP, the ending point endP, the control point controlP1 and the control point controlP2.

[0065] Preferably, step S4 specifically includes:

[0066] For the job performance overview view, job timing analysis view, and order process comparison view, two forms of visual interactive response methods, mouse clicking and mouse hovering, are constructed respectively.

[0067] The beneficial effects of the technical solution provided by the present invention are: combined with customized views and flexible interactions, it can assist warehouse managers in perceiving the operating status of picking operations from the overall to the details, analyze the time-varying patterns of picking operations, reveal the evolution of order status within picking batches, discover and analyze abnormal order processing processes, and thus provide support for warehouse managers in making production decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a schematic diagram of the overall process of the present invention.

[0069] Figure 2 It is a schematic diagram of the overall layout and visual mapping of the job performance overview view in the present invention.

[0070] Figure 3 It is a schematic diagram of the linear color mapping scheme of the annular calendar chart in the present invention.

[0071] Figure 4 In the figure, (a) is a schematic diagram of the visual mapping of the nested ring belts in the present invention, and (b) is a schematic diagram of the visual mapping of the nested radar chart in the present invention.

[0072] Figure 5 It is a schematic diagram of the overall layout and visual mapping of the job timing analysis view in the present invention.

[0073] Figure 6 It is a schematic diagram of the overall layout and visual mapping of the order process comparison view in the present invention.

[0074] Figure 7It is a schematic diagram of drawing a cubic Bezier curve in the present invention.

[0075] Figure 8 In the figure, (a) is a schematic diagram of the floating interaction of the annular calendar chart in the present invention; (b) is a schematic diagram of the floating interaction of the nested radar chart in the present invention; (c) is a schematic diagram of the click interaction of the selector in the present invention.

[0076] Fig. 9 In the figure, (a) is a schematic diagram of the floating interaction of the "strip" of the job timing analysis view in the present invention; (b) is a schematic diagram of the floating interaction of the stacked rectangles in the job timing analysis sub-graph of the present invention; (c) is a schematic diagram of the click interaction of the selector in the job timing analysis sub-graph of the present invention.

[0077] Fig.10 It is a visual mapping diagram of the picking batch details table in the present invention.

[0078] Fig.11 In the figure, (a) is an interactive design schematic diagram of a floating interactive schematic diagram of a flow chart in an order process comparison view in the present invention; (b) is a click interactive schematic diagram of a timeline zoom button in an order process comparison view in the present invention. DETAILED DESCRIPTION

[0079] Hereinafter, the technical solution provided by the present invention will be further elaborated in detail with reference to the accompanying drawings.

[0080] The overall process diagram of the process monitoring visualization method for e-commerce warehouse picking operations of the present invention is as follows: Figure 1 As shown, the specific steps are as follows:

[0081] Step 1: Data acquisition and processing

[0082] 1. Data acquisition

[0083] The outbound order record table is obtained from the warehouse management system of the warehousing service provider, which describes the process of the e-commerce warehouse creating picking information according to customer orders, assigning picking tasks, pickers picking goods according to tasks, and closing picking orders. Its specific attributes include: picking batch code, order code, product code, pickers, and timestamps of four types of operation events (i.e., creation, assignment, picking, and order closing), accurate to seconds. The creation timestamp indicates the time when the picking task is created, the assignment timestamp indicates the time when the product inventory is assigned and the picking batch is generated, the picking timestamp indicates the time when the picker receives the task and picks the goods, and the order closing time indicates the time when the picking ends and the order is closed.

[0084] 2. Data cleaning and conversion

[0085] This method processes erroneous records and missing fields in warehouse operation data through data cleaning, so as to optimize data quality and improve the accuracy of data analysis.

[0086] (1) Delete incorrect data

[0087] In the warehouse picking operation, the steps of creating, allocating, picking and closing orders in the picking batch are strictly carried out in chronological order. However, in the outbound order data, there are some records whose "closing time" is earlier than the "creation time", so delete these records with logical errors in time attributes.

[0088] (2) Delete missing data

[0089] There are two picking methods in warehouse picking operations: batch picking and individual picking. When the warehouse has a large number of orders, picking batches will be created to pick orders in batches to improve the efficiency of picking operations. This method focuses on the operation process supervision under order batch picking, so the fields with empty "picking batch code" attribute values ​​are deleted.

[0090] (3) Define performance indicators

[0091] According to the process of picking operation, four performance indicators for evaluating the operation status of picking operation are defined: total number of picking batches, total number of orders, total number of picked goods and total number of picked goods types.

[0092] (4) Calculate picking operation performance overview data

[0093] The time attribute in the outbound order data is converted from the original seconds to the "year-month-day" format, and the total number of picking batches, the total number of orders, the total number of picked goods, and the total number of picked goods types in the warehouse are counted on a daily basis to obtain an overview of the picking operation performance.

[0094] (5) Calculate the timing analysis data of picking operations

[0095] The time attribute in the outbound order data is converted from the original seconds to the "year-month-day-hour" format, and the total number of picking batches in the creation, allocation, picking and closing stages in the warehouse every hour every day is counted with the hour as the time granularity to obtain the picking operation timing analysis data.

[0096] (6) Calculate order processing data

[0097] Taking the picking batch as the analysis unit and setting multiple time slices, the overall processing time of each picking batch, the time slice division unit, the number of time slice divisions, and the order status list corresponding to each time slice are calculated to obtain the order processing process data.

[0098] 3. Data storage

[0099] After completing the above data cleaning and conversion operations, the picking operation performance overview data, picking operation timing analysis data and order processing process data are stored in the MySQL database for subsequent visualization.

[0100] Optionally, the MySQL database can be replaced by a PostgreSQL database, a TiDB database, or other databases.

[0101] Step 2: Visual Mapping

[0102] After data acquisition and processing, the operation performance overview view (such as Figure 2 As shown in the figure), job timing analysis view (as shown in the figure), Figure 3 ) and the order flow comparison view (as shown Figure 4 As shown in the figure, a visual mapping scheme is designed.

[0103] 1. Job performance overview view

[0104] Color: The color of the dots in the circular calendar chart reflects the specific value of a certain performance indicator on that day. This method uses four colors and the gradient between them to reflect the values ​​of the four job performance indicators. For example, dark blue is used to represent the minimum value (Min) of the indicator, light blue is used to represent the lower quartile (Q1) of the indicator; light red is used to represent the upper quartile (Q2) of the indicator, and dark red is used to represent the maximum value (Max) of the indicator. A gradient color system of four colors is used for linear mapping between the minimum value, the lower quartile, the upper quartile, and the maximum value.

[0105] Position: In the nested ring belt, the sub-ring belt obtained by rotating clockwise from 12 o'clock as the starting point reflects the statistical data (including maximum, minimum and median) of various operation performance indicators in different months. In the nested radar chart, the radial scale axis obtained by rotating clockwise from 12 o'clock as the starting point reflects the average value of various operation performance indicators in different months.

[0106] Area: In the nested bands, the area enclosed by the inner and outer boundary lines of each sub-band reflects the maximum value of the selected indicator in the month. The larger the area, the greater the fluctuation range of the indicator in the month. In the nested radar chart, the length of the vertex from the center point on the radial scale axis is used to encode the average value of the performance indicator.

[0107] 2. Job Timing Analysis View

[0108] Shapes: Use rectangles to represent various events in a picking operation.

[0109] Color: Use any four colors min 、ColorQ1 、Color Q3 、Color max , such as purple, blue, pink and orange, which respectively represent four events (creation, allocation, picking and order closing).

[0110] Position: Use different positions of the job timing analysis view to represent different time periods of the day. Starting from scale 0, 24 columns of stacked rectangles represent each hour from 0 to 24 o'clock from left to right. Each column of stacked rectangles (job timing analysis subgraph) is arranged in the order of creation, allocation, picking, and closing, and is distinguished by a certain space interval. In the same job timing analysis view, rectangles in the same event but different time periods will be connected by "strips" of the same color, and the transparency of the "strips" will be appropriately reduced for distinction.

[0111] Height: The height of each rectangle represents the number of picking batches in a certain event. The higher the height, the more picking batches there are. The total height of each column of the job timing analysis subgraph represents the total number of picking batches in the four events in a certain period of time.

[0112] 3. Order process comparison view

[0113] Shape: Use any four plane graphics, such as triangle, circle, square and diamond event symbols to map the four event states of an order in a picking batch: creation, allocation, picking and order closing.

[0114] Color: Use any four colors, such as purple, blue, pink and orange to represent the four events (creation, allocation, picking and closing) respectively.

[0115] Size: Use the size of the event symbol to encode the number of orders processed in different time slices. For example, when the number of orders created in the second time slice exceeds the number of orders created in the first time slice, the size of the triangle symbol corresponding to the second time slice scale is larger than the first one.

[0116] Position: The timeline is placed horizontally, and the event symbols are located above the timeline. Event symbols in the same time slice are connected by straight lines and arranged vertically in the order of events. The end symbols and beginning symbols of adjacent time slices are connected using cubic Bezier curves. Event symbols correspond to the time slice scales.

[0117] Optionally, any two of the “four colors” involved in the above job performance overview view, job timing analysis view and order process comparison view are not the same.

[0118] Step 3: Visualization layout and implementation

[0119] Read the outbound order data of the specified year and month, and complete the layout and implementation of the outer ring calendar chart, inner nested ring belt, and nested radar chart in the job performance overview view; complete the layout and implementation of the upper text label, legend, and control panel, and the lower stacked chart in the job timing analysis view; complete the layout and implementation of the upper text label, legend, and control panel, and the lower flow chart in the order process comparison view.

[0120] The details are as follows:

[0121] 1. Visualization layout and implementation of job performance overview view

[0122] The overall layout of the job performance overview view is as follows: Figure 2 As shown. The selector on the upper left is used to switch between different job performance indicators, and the color legend on the upper right uses four colors and gradient colors between them to reflect the values ​​of the four job performance indicators. This method uses dark blue to represent the minimum value (Min) of the indicator, light blue to represent the lower quartile (Q1) of the indicator; light red to represent the upper quartile (Q2) of the indicator, and dark red to represent the maximum value (Max) of the indicator. A gradient color system of four colors is used for linear mapping between the minimum value, lower quartile, upper quartile, and maximum value to accurately reflect the changes in the values ​​of each indicator.

[0123] The mapping scheme of the dot fill color in the circular calendar chart is as follows Figure 3 As shown, calculate the dot fill color Hex i The method is as follows:

[0124]

[0125] Where: Hex i ——The hexadecimal code of the color of the performance indicator value; Value i ——Specific value of performance indicator; Value min ——Minimum value of performance indicator; Value max ——Maximum value of performance index; Hex max ——The hexadecimal encoding of the color of the maximum value of the performance indicator; Hex min ——The hexadecimal code of the color of the minimum value of the performance indicator.

[0126] Based on the predefined canvas boundaries, the absolute coordinates of each dot in the circular calendar chart are calculated; for the calculated absolute coordinates, the dots are drawn in sequence and filled with the colors mapped to the corresponding performance indicator values. The method for calculating the dot coordinates is as follows:

[0127]

[0128]

[0129]

[0130] Where: θ radians ——radians, indicating the angle that each month needs to be rotated in the canvas; i——month, i=1,2,...,N,N∈[1,2,3,...,12]; x local ,y local ——Relative coordinates of the dot; col——The day of the week corresponding to the date of the dot (counting starts from 0, 0 represents Sunday, 6 represents Saturday); row——The week number of the month corresponding to the date of the dot (counting starts from 0); Radius——The radius of the dot; circleGap——The spacing between the dots; verticalPos——The vertical distance between the dot and the top of the canvas; x global ,y glocal ——the absolute coordinates of the dot; s——the canvas scaling factor (0 <s<1)。

[0131] The layout design of the nested belt is as follows Figure 4 As shown in (a), the central axis of the first sub-ring is located at 12 o'clock, and then five sub-rings are obtained by rotating in a clockwise direction by an angle θ. Each sub-ring represents the statistical data of one month, and the sub-rings are connected from beginning to end to form a closed interval. The dotted background of the ring represents the radial scale, and the innermost ring circle inner Indicates the minimum value of the statistical data in five months Month min , the outermost circle Circle outer Indicates the maximum value of the statistical data in five months Month max ; and the line between the outermost and innermost circles represents the average value of the five-month statistics Month median In each sub-zone, inner and outer boundary lines are used to represent statistical values.

[0132] The layout design of the nested radar chart is as follows Figure 4 As shown in (b), with point O as the rotation center and 12 o'clock as the starting position, five radial scale axes are obtained by rotating in a clockwise direction by an angle θ to represent five months. With point O as the diffusion center, three dotted concentric circles are nested. The first concentric circle Ring first Indicates the minimum value of the average value in five months, the third concentric circle Ring third Indicates the maximum value of the average value in five months. The lines connecting the center point O to the vertices on the five scale axes encode the average values ​​of the performance indicators for five months. For example, when the selected indicator is the number of picking batches, Figure 4 In (b), the length of OA encodes the average number of picked batches in January, and the length of OE encodes the average number of picked batches in May.

[0133] 2. Visualization layout and implementation of job timing analysis view

[0134] The overall layout of the job timing analysis view is as follows: Figure 5 As shown. On the upper left are text labels and legend labels. The text labels indicate the date corresponding to the current stacking chart, and the legend labels indicate the corresponding relationship between the stacking rectangles and the four types of events. On the upper right are the stacking order selector and the view order adjustment button group. On the lower side is a stacking chart that reflects the timing rules of the picking operation.

[0135] The horizontal axis of the stacked chart uses a strip scale to map integers from 0 to 23 to the specified canvas width. Each strip represents one hour, and the strip width (bandwidth) represents the width of the stacked rectangle. The vertical axis uses a linear scale to map the number of batches to the specified canvas height. The coordinates of the starting point of the stacked rectangle in each column are calculated as follows:

[0136]

[0137] Where: rect x ,rect y ——The horizontal and vertical coordinates of the drawing starting point; StackSvgWidth, StacksvgHeight——constants, view canvas width and height; i——time; stackCnt——number of stacked batches; initCnt——number of original batches; initCnt——strip interval; BathSum——canvas batch capacity; rectGap——rectangular interval.

[0138] 3. Visual layout and implementation of order process comparison view

[0139] The overall layout of the order process comparison view is as follows Figure 6 As shown. On the upper left are text labels and legend labels. The text labels indicate the date corresponding to the current flow chart, and the legend labels indicate the correspondence between event symbols and four types of events. On the upper right are the view order adjustment button group and the timeline zoom button. On the lower side is a flow chart reflecting the evolution of order status.

[0140] 1) Event symbol mapping

[0141] According to the event type processType of the order, the event symbol stateSymbol is calculated by the symbol generation function symbolGenerator:

[0142] stateSymbol processType =symbolGenerator(processType)

[0143] In the formula: processType - four events: create order, assign order, pick order or close order; stateSymbol - triangle, circle, square or diamond corresponding to the event.

[0144] Since the number of orders in different time slices varies greatly, in order to ensure that different order quantities can be clearly compared within the same view range, the symbol size is mapped by the logarithmic scale in D3.js to handle the large range of data changes. The domain of this scale is the minimum and maximum value of the order quantity in different time slices, and the value range is a custom pixel range.

[0145] The X-axis coordinate of the time axis is determined by the linear scale calculation. The domain of the linear scale is determined by the number of time slices timeSliceNum corresponding to the picking batch processing process, and the range is the canvas width. Orders are divided according to the time attribute of the picking operation data, and orders belonging to the same time slice will be attributed to one time scale. The Y-axis coordinate is determined by the number of symbols in the same time slice, and its calculation method is as follows:

[0146] symbol y =startY+(i-1)×gapSpace

[0147] Where: startY is the starting position of the first event symbol in the i-th time slice; gapSpace is the spatial interval of event symbols.

[0148] 2) Event symbol connection drawing

[0149] In the design of event symbol connection, symbols belonging to the same time slice are connected by straight lines, and symbols of adjacent time slices are connected by cubic Bezier curves. Figure 7 As shown, four points are needed to draw a cubic Bezier curve, namely: the starting point startP, the end point endP, and two control points controlP1 and controlP2. The starting point and the end point determine the starting and ending positions of the curve, and the two control points determine the shape and direction of the curve. In actual design, the positions of the four points need to be manually set as needed to obtain a curve shape that meets the requirements. In this method, the calculation method of the two control points is as follows:

[0150]

[0151]

[0152] Where: startPX i ——X-axis coordinate of the starting point of the i-th time slice; startPY i——Y-axis coordinate of the starting point of the i-th time slice; endPX i ——The X-axis coordinate of the end point of the i-th time slice; endPY i ——Y-axis coordinate of the end point of the i-th time slice; controlVal——the offset distance of the control point in the horizontal position.

[0153] Step 4: Interaction Design

[0154] 1. Interactive design of job performance overview view

[0155] The job performance overview view supports two mouse interactive operations: hovering and clicking:

[0156] Suspension: Figure 8 As shown in (a), when the mouse hovers over the dot on the circular calendar chart, a prompt box will appear, showing the specific information of the date and the number of picking batches. When the mouse is moved away, the prompt box will automatically hide. Figure 8 As shown in (b), when the mouse hovers over a node in the nested radar chart, a tooltip box will appear to display the mean number of picking batches.

[0157] Click: Figure 8 As shown in (c), you can switch to view different performance indicators by clicking the selector above the view. The data of the circular calendar chart, nested ring belt, and nested radar chart will be updated synchronously. Click any dot with the mouse to add a shadow effect to the dot; at the same time, the detailed operation status of the corresponding date is displayed in the job sequence analysis view. Click the dot with the mouse again, the shadow effect disappears, and the detailed operation status of the corresponding date is removed from the job sequence analysis view.

[0158] 2. Interactive design of job timing analysis view

[0159] The job timing analysis view supports two mouse interactive operations: hovering and clicking:

[0160] Suspension: Fig. 9 As shown in (a), in the stacked graph (job timing analysis view), if you hover the mouse over any color "strip", the stacked rectangle (job timing analysis subgraph) and "strip" of that color will be highlighted, and the transparency of the stacked rectangles and "strips" of other colors will be reduced. At the same time, the stacked rectangles and "strips" of other juxtaposed stacked graphs will also show the same changes to support vertical comparison. Fig. 9 As shown in (b), when the mouse hovers over any stacked rectangle, a tooltip box will appear, showing the number of picking batches in a specific event.

[0161] Click: By default, the original stacking order of each column of stacked rectangles in the stacked chart is from top to bottom: creation - allocation - picking - order closing. To ensure the accuracy of data comparison, click the selector in the upper right corner of the stacked chart to switch the order of stacked rectangles to view the alignment of stacked rectangles of various events. Fig. 9 As shown in (c), after displaying the "picking" event as the first item in the stacking order, the order of the stacking rectangles will be adjusted to: picking - closing the order - creating - allocating. When there are multiple stacking charts placed side by side, you can switch the arrangement order of the stacking charts by clicking the button group on the upper right side of the stacking chart. The button group contains four function buttons from left to right: the first button places the current stacking chart at the top of the canvas; the second button moves the current stacking chart up one position; the third button places the current stacking chart at the bottom of the canvas; the fourth button moves the current stacking chart down one position. Click any stacking rectangle with the mouse, and a table will pop up above the view to display all the picking batch information in the specific event in the current time period. As shown Fig.10 As shown in the figure, each row in the table represents a picking batch and shows the detailed information of the batch. The six columns from left to right are: batch code batch, order quantity O.Qty, product quantity P.Qty, product type number P.Type, batch processing time Process and optional operation Operation. Fig.10 As shown in area a, columns O.Qty, P.Qty, P.Type, and Process all implement sorting functions and can be arranged in ascending or descending order. Fig.10 As shown in area b, in the Operation column, two function buttons ViewBatch and DeleteBatch are set: Click the ViewBatch button, as shown in Fig.10 As shown in area c, the current row is highlighted, and the order process analysis view will update the order processing process of the corresponding picking batch in this row; click the DeleteBatch button to cancel the highlight of the current row, and the order processing of the corresponding picking batch will be removed from the order process analysis view. The content of the batch column is designed to be hidden by default. When the mouse hovers over the column, a prompt box pops up to display the complete batch code, such as Fig.10 As shown in area d in the figure. Click the close button in the upper right corner. Fig.10 As shown in area e in the figure, the current table is removed from the view.

[0162] 3. Interactive design of job timing analysis view

[0163] The job timing analysis view supports hover and click mouse interactive operations:

[0164] Hover: When the mouse hovers over any event symbol, a prompt box will appear to display the time corresponding to the current event symbol (in the form of n*timeSliceUnit, where n represents the number of time slices and timeSliceUnit represents the time slice granularity), as well as the corresponding order quantity. Fig.11 As shown in (a), when the mouse hovers over the second triangle, a prompt box appears showing that the current processing time is 2*3600, or 7200 seconds. 11 orders are created in the second time slice.

[0165] Click: When there are multiple juxtaposed flow charts, you can switch the arrangement order of the flow charts by clicking the button group in the upper right corner of the flow chart with your mouse. The functions of each button are consistent with the job timing analysis view. You can flexibly adjust the granularity of the time slice by clicking the zoom in / out button in the upper right corner of the flow chart with your mouse. Click the "-" button to increase the span of the time slice (for example, from every 300 seconds to every 600 seconds), thereby shortening the length of the time axis. Click the "+" button to reduce the span of the time slice (for example, from every 600 seconds to every 300 seconds), thereby extending the length of the time axis. Fig.11 (a) and Fig.11 As shown in (b), the event symbol distribution of the same picking batch at different time slice granularities (every 3600 seconds and every 4800 seconds) is different.

[0166] The technical solution provided by the present invention can be combined with customized views and flexible interactions, which can assist warehouse managers in perceiving the operating status of picking operations from the overall to the details, analyzing the time-varying patterns of picking operations, revealing the evolution of order status within picking batches, discovering and analyzing abnormal order processing processes, and thus providing support for warehouse managers in making production decisions.

Claims

1. A process monitoring visualization method for picking operations in e-commerce warehouses. It is characterized in that The following steps are involved: Step S1: Acquire manual picking process data, and store the cleaned and converted manual picking process data into a database; Step S2: Visually map the manual picking process data through the visual channel, including: constructing an operation performance overview view, an operation timing analysis view, and an order process comparison view; Step S3: Divide the canvas into multiple areas, and arrange the job performance overview view, job timing analysis view, and order process comparison view respectively; Step S4: Based on the job performance overview view, the job timing analysis view and the order process comparison view, a visual interactive response method for viewing the manual picking process data is constructed.

2. A process monitoring visualization method for picking operations in e-commerce warehouses as claimed in claim 1, It is characterized in that Step S1 specifically includes: Step S11: Obtaining a delivery order record table from the warehouse management system of the warehousing service provider; The attributes of the outbound order record table include: picking batch code, order code, product code, picking personnel, creation timestamp, allocation timestamp, picking timestamp and order closing timestamp; Step S12: data cleaning and data conversion; Step S121: deleting erroneous data and missing data; Step S122: convert the time attribute in the outbound order data from seconds to "year-month-day"; Step S123: Counting four performance indicators of the warehouse's daily picking operation, including: the total number of picking batches, the total number of orders, the total number of picked goods, and the total number of picked goods types, to obtain operation performance overview data; Step S124: Count the total number of picking batches in the four events of creation, allocation, picking and order closing that leave the warehouse every hour every day to obtain operation timing analysis data; Step S125: Taking the picking batch as the analysis unit, setting multiple time slices, calculating the overall processing time of each picking batch, the time slice division unit, the number of time slice divisions, and the order status list corresponding to each time slice, and obtaining the order processing process data; Step S13: Store the operation performance overview data, operation timing analysis data and order processing process data in the database.

3. A process monitoring visualization method for picking operations in e-commerce warehouses as claimed in claim 2, It is characterized in that The database is a MySQL database.

4. A process monitoring visualization method for picking operations in e-commerce warehouses as claimed in claim 2, It is characterized in that (1) Building a job performance overview view based on job performance overview data; For each performance indicator, four colors are used to represent the minimum value, lower quartile, upper quartile and maximum value of a single day in the circular calendar chart; the values ​​of the corresponding performance indicators are linearly mapped between the four colors through a gradient color system, and each dot representing a single day is filled with a different color; For each performance indicator, the nested ring is divided into multiple sub-rings according to the number of months, and the outer edge, inner edge and middle line of the sub-ring are used to represent the maximum value, minimum value and median of the corresponding month respectively; in the nested radar chart, the radial scale axis represents the average value of the corresponding month; (2) Building a job timing analysis view based on the job timing analysis data; Rectangles of four colors represent creation, allocation, picking, and order closing in a single hour, and different lengths represent different quantities. The rectangles are stacked into a job timing analysis subgraph in the order of creation, allocation, picking, and order closing in units of hours, and arranged in chronological order. Connect the rectangles of the same color in each job timing analysis sub-graph in time sequence in the form of strips; (3) Construct order process comparison view based on order processing data; Four plane graphics are used to represent creation, allocation, picking and order closing, and the size of the plane graphics represents different quantities; On the time coordinate axis, events in the same time slice are connected by straight lines and arranged vertically in order; adjacent plane figures on adjacent time slices are connected by cubic Bezier curves.

5. A process monitoring visualization method for picking operations in e-commerce warehouses as claimed in claim 4, It is characterized in that (1) The overview view of the layout job performance includes: Calculate the fill color of the dots, In the formula, Hex i is the hexadecimal code of the color corresponding to the i-th performance indicator value; Value i is the i-th performance indicator value, Value min 、Value max They are the minimum and maximum values ​​of the performance indicators, Hex max 、Hex min They are the hexadecimal codes corresponding to the maximum and minimum values ​​of the performance indicators respectively; Based on the predefined canvas boundaries, calculate the absolute coordinates of each point in the circular calendar chart. In the formula, θ radians is the angle that each month needs to be rotated in the canvas; j represents the month, j=1,2,...,N, N∈[1,2,3,...,12]; (x local ,y local ) is the relative coordinate of the dot; col represents the date number of the dot corresponding to the day of the week, col = 0, 1, 2, ... 6; row represents the week number of the dot corresponding to the day of the month, row = 0, 1, 2, ...; Radius is the radius of the dot, circleGap is the distance between the dots, verticalPos is the vertical distance between the dot and the top of the canvas, (x global ,y glocal ) is the absolute coordinate of the dot; s is the canvas scaling factor, 0 <s<1; Draw dots according to their absolute coordinates and fill them with colors; In the center of the annular calendar chart composed of multiple months, nested ring belts and nested radar charts are arranged from the outside to the inside; Among them, multiple sub-rings are connected end to end to form a closed interval, and the dotted background of the ring represents the radial scale; the nested radar chart includes three dotted concentric circles, from the inside to the outside, the first concentric circle represents the minimum value of the average value of multiple months, and the third concentric circle represents the maximum value of the average value of multiple months; (2) The job arrangement timing analysis view includes: Calculate the starting point for drawing each job timing analysis subgraph, In the formula, (rect x ,rect y ) is the coordinate of the drawing starting point, StackSvgWidth and StackSvgHeight are the width and height of the canvas respectively, bandGap is the interval of the band, k is the hour number, stackCnt is the number of stacked batches, initCnt is the number of original batches, maxBathSum is the maximum batch capacity of the canvas, and rectGap is the interval of the job timing analysis subgraph; Draw each job timing analysis subgraph based on the drawing starting point; (3) Arrange the order process comparison view including: Determine the plane figure corresponding to the event, stateSymbol processType =symbolGenerator(processType); In the formula, processType is the event type, stateSymbol processType is the plane figure corresponding to the processType event, and symbolGenerator(·) is the generating function of the plane figure; The domain of the linear scale is determined by the number of time slices corresponding to the picking batch processing process, and the range of the linear scale is determined by the canvas width; the X-axis coordinate of the time coordinate axis is determined by the linear scale; Determine the Y-axis coordinate symbol of the time axis y , symbol y =startY+(l-1)×gapSpace; Where startY is the starting position of the first event symbol in the lth time slice, and gapSpace is the spatial interval between the plane figures; Calculate control point controlP1 and control point controlP2, Where startPX l is the X-axis coordinate of the starting point of the lth time slice, startPY l ——The Y-axis coordinate of the starting point of the lth time slice, endPX l is the X-axis coordinate of the end point of the lth time slice, endPY l is the Y-axis coordinate of the end point of the lth time slice, and controlVal is the offset distance of the control point in the horizontal position; Draw a cubic Bezier curve based on the starting point startP, the ending point endP, the control point controlP1 and the control point controlP2.

6. A process monitoring visualization method for e-commerce warehouse picking operations as claimed in claim 1, It is characterized in that Step S4 specifically includes: For the job performance overview view, job timing analysis view, and order process comparison view, two forms of visual interactive response methods, mouse clicking and mouse hovering, are constructed respectively.