Inspection system, inspection method, and program

Abstract

This invention provides an inspection system, inspection method, and program that are advantageous for determining multiple control parameters that may affect the transport of electronic components. [Solution] The control unit of the inspection system acquires transport status information including a plurality of interrelated control parameters and cumulative data regarding transport results (S1), derives a parameter set that includes a plurality of control parameters and shows the smallest uninspected rate from the transport status information using Gaussian process regression (S2, S3), and updates a plurality of control parameters for the transport device and peripheral equipment respectively based on the derived parameter set (S4).

Description

[Technical Field]

[0001] This disclosure relates to inspection systems, inspection methods, and programs. [Background technology]

[0002] Electronic components such as capacitors (e.g., multilayer ceramic capacitors) can be transported by a transport device such as a transport table while undergoing various inspections (see, for example, Patent Document 1). [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2019-86397 [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] If electronic components being transported by a transport device are not in the target transport state, it may be effective to adjust the control parameters of multiple devices included in the inspection system. When adjusting the control parameters of multiple devices, for example, a user may manually and repeatedly adjust a specific control parameter based on their intuition, while temporarily fixing the control parameters other than that specific one.

[0005] However, when adjusting multiple control parameters one by one, the number of adjustment steps tends to increase, the time required for adjustment becomes longer, and this can ultimately lead to a decrease in the equipment's production capacity. Furthermore, since the adjustment of control parameters is left to the judgment and experience of the adjuster, the effort required for adjustment and the accuracy of the adjustment tend to vary among adjusters.

[0006] This disclosure provides a technique advantageous for determining multiple control parameters that may affect the transport of electronic components. [Means for solving the problem]

[0007] One aspect of the present disclosure relates to an inspection system comprising: a transport device for transporting a plurality of electronic components; an inspection device for inspecting the plurality of electronic components being transported by the transport device; peripheral equipment for adjusting state variables affecting the transport of the plurality of electronic components; and a control unit for acquiring transport results including a supply rate indicating the number of electronic components transported per unit time by the transport device and an uninspected rate indicating the percentage of electronic components among the plurality of electronic components that were not inspected due to transport failures, wherein the control unit controls the operation of the transport device and peripheral equipment by adjusting a plurality of control parameters relating to the transport device and peripheral equipment, the control unit acquires transport state information including a plurality of interrelated control parameters and cumulative data relating to the transport results, derives a parameter set including a plurality of control parameters that shows the smallest uninspected rate from the transport state information using Gaussian process regression, and updates a plurality of control parameters relating to the transport device and peripheral equipment based on the derived parameter set.

[0008] The peripheral device includes a first peripheral device having a charging section that electrostatically attracts electronic components to a transport device with an attractive force corresponding to the applied voltage, and a second peripheral device having a vacuum suction section that attracts electronic components with an attractive force corresponding to the airflow rate. The parameter set may include control parameters related to the applied voltage to the charging section and control parameters related to the airflow rate in the vacuum suction section.

[0009] The control unit may use Gaussian process regression to estimate multiple transport states from transport state information, including combination data of multiple control parameters selected from the search range of multiple control parameters and the uninspected rate associated with the multiple control parameters. Among the multiple estimated transport states, the control parameters of the transport state in which the uninspected rate is minimized under the condition that the supply speed is equal to or greater than the set speed may be used as multiple control parameters included in the parameter set.

[0010] The control unit may acquire the latest multiple control parameters and the latest transport results after the update, and derive a parameter set from transport status information that includes cumulative data reflecting the latest multiple control parameters and the latest transport results.

[0011] The state variables affecting the transport of multiple electronic components may include one or more of the following: the rotational speed of the transport table on which the transport device has and on which the multiple electronic components are placed; the airflow rate in the vacuum suction section that attracts each of the multiple electronic components when they are placed on the transport table; the voltage applied to the charged section that electrostatically attracts each of the multiple electronic components to the transport table; the humidity around the transport table; the temperature around the transport table; the drive parameters of the component supply device that supplies the multiple electronic components to the transport table; the amount of clearance between the component supply device and the transport table; and the type of alignment guide that aligns each of the multiple electronic components on the transport table.

[0012] Another aspect of this disclosure relates to an inspection method comprising the steps of transporting a plurality of electronic components by a transport device, inspecting the plurality of electronic components being transported by the transport device by an inspection device, obtaining transport results including a supply rate indicating the number of electronic components transported per unit time by the transport device and an uninspected rate indicating the percentage of electronic components among the plurality of electronic components that were not inspected due to transport defects, and adjusting a plurality of control parameters for each of the peripheral device and the transport device to adjust state quantities affecting the transport of the plurality of electronic components, wherein the step of adjusting a plurality of control parameters to control the operation of the transport device and the peripheral device is to obtain transport state information including a plurality of interrelated control parameters and cumulative data relating to the transport results, deriving a parameter set including a plurality of control parameters that shows the smallest uninspected rate from the transport state information using Gaussian process regression, and updating a plurality of control parameters for each of the transport device and the peripheral device based on the derived parameter set.

[0013] Another aspect of the present disclosure is a program for causing a computer to execute procedures for transporting a plurality of electronic components by a transport device, procedures for inspecting the plurality of electronic components being transported by the transport device by an inspection device, a supply rate indicating the number of electronic components transported per unit time by the transport device, and an uninspected rate indicating the ratio of electronic components for which inspection was not performed due to transport failures among the plurality of electronic components, and obtaining a transport result including these, and adjusting a plurality of control parameters related to each of the peripheral device and the transport device that adjust a state quantity affecting the transport of the plurality of electronic components to control the operations of the transport device and the peripheral device. In the procedure for adjusting a plurality of control parameters to control the operations of the transport device and the peripheral device, transport state information including a plurality of control parameters related to each other and cumulative data related to the transport result is obtained, and from the transport state information, a parameter set including a plurality of control parameters and indicating the minimum uninspected rate is derived using Gaussian process regression, and based on the derived parameter set, a plurality of control parameters related to each of the transport device and the peripheral device are updated. This relates to the program.

Effects of the Invention

[0014] According to the present disclosure, it is advantageous for determining a plurality of control parameters that can affect the transport of electronic components.

Brief Description of the Drawings

[0015] [Figure 1] FIG. 1 is a diagram showing a schematic configuration of an example of an inspection system. [Figure 2] FIG. 2 is a cross-sectional view showing an example of a vacuum suction portion provided in an alignment guide. [Figure 3] FIG. 3 is a functional block diagram showing an example of a control configuration of an inspection system. [Figure 4] FIG. 4 is a flowchart showing an example of a transport control flow.

Modes for Carrying Out the Invention

[0016] Hereinafter, embodiments of the disclosed technology will be described with reference to the drawings.

[0017] In the following description, unless otherwise specified, the terms "upstream" and "downstream" are based on the conveyance of the work (electronic component).

[0018] FIG. 1 is a diagram showing a schematic configuration of an example of the inspection system 100.

[0019] The inspection system 100 shown in FIG. 1 includes a work supply device 1 that supplies a work to a conveyance device 2 (including a conveyance table 2a and a conveyance drive device 2b). The work supply device 1 of the present embodiment includes a parts feeder (not shown) that can store a large number of works and sends out the stored works downstream, and a linear feeder 1a that further sends out the works sent from the parts feeder downstream and places them on the conveyance table 2a. The linear feeder 1a sends out each work so as to place it at the transfer point on the upper surface of the conveyance table 2a. The work supply device 1 (parts feeder and linear feeder 1a) of this example sends out each work downstream by vibration. By controlling the drive voltage and drive frequency for the work supply device 1 and appropriately adjusting the amplitude and / or frequency of the vibration of the work supply device 1, it is possible to change the state of sending out each work by the work supply device 1.

[0020] The work to be inspected in the inspection system 100 is composed of a chip component (such as a capacitor or a resistor) or any other electronic component and can have any characteristics. The work of this example has a hexahedral shape (cuboid shape) and includes a main body made of an insulator and electrodes made of conductors provided at both ends in the longitudinal direction of the main body. The work is placed on a conveyance table 2a having a circular planar shape and is conveyed along an arc-shaped conveyance path 5 as the conveyance table 2a rotates (refer to the arrow indicated by "X" in FIG. 1).

[0021] As the workpieces are transported along the transport path 5, the imaging unit 20 captures images of the outer surfaces of each workpiece on the transport table 2a (six surfaces in this example: top, bottom, left, right, front, and back), thereby acquiring images of the external appearance of each workpiece (six images in this example). The imaging unit 20 shown in Figure 1 has a side camera unit 8, an inner camera unit 9, an upper camera unit 10, a lower camera unit 11, a front camera unit 12, and a rear camera unit 13, which are sequentially arranged downstream. The transport table 2a in this example is made of a transparent material (for example, glass), and the lower surface of each workpiece that contacts the upper surface of the transport table 2a is captured through the transparent transport table 2a by the imaging unit 20 located below the transport table 2a (the lower camera unit 11 in the example shown in Figure 1).

[0022] Under the control of the control unit 50, each workpiece is visually inspected based on the image analysis results of the captured images. Depending on the results of the visual inspection, each workpiece on the transport table 2a is collected into a corresponding storage box (not shown) by the discharge unit 14 located downstream of the imaging unit 20. The discharge unit 14 collects workpieces that have been found to be defective (e.g., scratches or chips) and workpieces that have not been found to be defective into separate storage boxes.

[0023] The inspection system 100 of this embodiment is further provided with a transfer and alignment unit 21 that transfers and aligns workpieces from the linear feeder 1a onto the transport table 2a, and a charging unit 6 that is located below the transport table 2a and functions as a holding means that charges the transport table 2a to hold the workpieces placed on the transport table 2a.

[0024] The transfer and alignment unit 21 includes a vibration-free section 4 provided between the linear feeder 1a and the transport table 2a, and an alignment guide 7 provided downstream of the vibration-free section 4.

[0025] Figure 2 is a cross-sectional view showing an example of vacuum suction sections 17 and 18 provided on the alignment guide 7. The workpiece W shown in Figure 2 is in the state immediately after it has been transferred from the vibration-free section 4 to the transfer point on the transport table 2a and transport by the rotation of the transport table 2a has begun.

[0026] The alignment guide 7 includes sides that constitute a guide surface 7a for aligning each workpiece W on the transport table 2a along the transport path 5 (see Figure 1), and this guide surface 7a has a straight shape when viewed from above. The workpieces W are electrostatically attracted to the upper surface of the transport table 2a, which is charged by the charging unit 6 (see Figure 1), and are transported along the guide surface 7a of the alignment guide 7.

[0027] A gap γ is formed between the lower surface 7z of the alignment guide 7 and the upper surface 2s of the transport table 2a. Meanwhile, a suction passage 18 is provided to pass through the inside of the alignment guide 7, with one end 18a of the suction passage 18 opening into the gap γ, and a suction device 17 connected to the other end 18b of the suction passage 18 via a relay line 17a.

[0028] As shown in Figure 2, the gap γ between the alignment guide 7 and the transport table 2a is constantly vacuum-suctioned by the suction device 17 and the suction passage 18 in the directions of arrows δ1, δ2, and δ3. This vacuum suction causes each workpiece W on the transport table 2a to be drawn toward the guide surface 7a and accurately positioned at the desired location (on the transport path 5 shown in Figure 1) by contacting the guide surface 7a. Furthermore, this vacuum suction prevents air from being ejected from the gap γ toward the workpiece W even if the dimensions of the gap γ change when the transport table 2a rotates (see the dotted line in Figure 2), thereby stably maintaining the state in which the workpiece W is electrostatically attracted to the desired location on the transport table 2a.

[0029] Figure 3 is a functional block diagram showing an example of the control configuration of the inspection system 100. Each functional block shown in Figure 3 can be implemented by any hardware and / or software, and may be implemented by a single device, by a combination of multiple devices, or by a single device implementing two or more functional blocks.

[0030] As shown in Figure 3, the inspection system 100 includes a control unit 50, a measuring device 51, a user interface 52, various instruments 55, and an external storage unit 56.

[0031] The measuring device 51 is a functional block that comprehensively represents equipment such as sensors for detecting various state quantities, and transmits the detected state quantities to the control unit 50. In this example, the measuring device 51 includes a hygrometer for detecting humidity around the transport table 2a and a thermometer for detecting temperature around the transport table 2a, and the detection results of humidity and temperature around the transport table 2a are provided to the control unit 50 from the hygrometer and thermometer.

[0032] The user interface 52 functions as an input device that accepts various types of information from the user (coordinator), and also functions as an output device that outputs various types of information to the user.

[0033] The various devices 55 comprehensively represent devices that receive control processing from the control unit 50. For example, the work supply device 1 shown in Figure 1 above, the transport drive device 2b that rotates the transport table 2a, the transfer and alignment unit 21, the imaging unit 20, the discharge unit 14, and the charging unit 6 may be included in the various devices 55. In this embodiment, the transport device 2 (transport table 2a) that transports multiple workpieces W, the inspection device 20 that inspects the multiple workpieces W being transported by the transport device 2, and peripheral devices that adjust state quantities affecting the transport of multiple workpieces W (for example, the work supply device 1, the charging unit 6, the vacuum suction units 17 and 18 provided together with the alignment guide 7, a humidifier / dehumidifier, a room temperature control device such as an air conditioner, etc.) are included in the various devices 55.

[0034] The external storage unit 56 is a storage unit provided outside the control unit 50, and appropriately receives various data recording processes and various stored data reading processes from the control unit 50.

[0035] The control unit 50 shown in Figure 3 includes an image analysis unit 60, a timer device 61, and an internal storage unit 62. The image analysis unit 60 can detect (inspect) whether a workpiece has cosmetic defects by analyzing the image of the workpiece acquired by the imaging unit 20. The timer device 61 can measure various times. The internal storage unit 62 records various data, and the stored data is read out as appropriate.

[0036] The timer device 61 may be provided as an external device to the control unit 50. The user interface 52 may be provided integrally with the control unit 50, or it may be provided separately from the control unit 50.

[0037] [Conveyor control flow] Next, we will describe an example of the workpiece transport control flow in the inspection system 100.

[0038] Figure 4 is a flowchart showing an example of a transport control flow.

[0039] In this embodiment, the control unit 50 controls the transport of multiple workpieces according to the transport control flow shown in Figure 4 while the inspection system 100 is in operation, while the transport control flow shown in Figure 4 is also interrupted when the inspection system 100 is stopped.

[0040] The control unit 50 acquires the transport status under the current transport conditions (S1 in Figure 4). That is, with the various devices of the inspection system 100 in operation, the control unit 50 acquires the transport status for the most recent t seconds (where "t" is an arbitrary natural number (e.g., t=60)) under the current transport conditions (control parameters) and records it in the internal storage unit of the control unit 50 (see internal storage unit 62 in Figure 3) and / or in an external storage unit accessible to the control unit 50 (see external storage unit 56 in Figure 3).

[0041] Here, "transportation conditions" refers to a set (combination) of multiple control parameters relating to the various devices 55 of the inspection system 100 (in particular, various devices that adjust state quantities affecting workpiece transport (including transport device 2 and peripheral devices)). The multiple control parameters included in the transport conditions may be parameters that can be set actively, or parameters that can be measured by a measuring device 51 (see Figure 3). Parameters that can be set actively may be obtained by the control unit 50 from data stored in the storage unit as setting parameter information (control parameters). Parameters that can be measured can be obtained by the control unit 50 when the measurement results of the corresponding measuring device 51 are received.

[0042] The "state quantities affecting workpiece transport" are not limited and include, for example, one or more of the following: the rotational speed of the transport table 2a on which multiple workpieces are placed; the airflow rate (supplied air suction amount) in the vacuum suction units 17 and 18 that suck each workpiece when it is placed on the transport table 2a; the voltage applied to the charging unit 6 that electrostatically attracts each workpiece to the transport table 2a (electrostatic attraction voltage); the humidity around the transport table 2a; the temperature around the transport table 2a; the drive parameters of the workpiece supply device 1 that supplies multiple workpieces to the transport table 2a (drive voltage and drive frequency (e.g., PF (parts feeder) voltage, PF frequency, LF (linear feeder) voltage, LF frequency)); the amount of clearance between the workpiece supply device 1 and the transport table 2a; and the type of alignment guide 7 that aligns each of the multiple electronic components on the transport table 2a. The "type of alignment guide 7" here can be determined, for example, by the shape, material and surface treatment of the alignment guide 7. For example, the alignment guide 7 may have a surface that contacts the workpiece with dimensions of 1.0 mm vertically and 40 mm horizontally, the material of the alignment guide 7 may be non-magnetic superhard steel, and the surface of the alignment guide 7 may be coated with DLC (diamond-like carbon). The "type of alignment guide 7" may be determined based on these characteristics.

[0043] For example, state variables whose actual values ​​(state variables) do not necessarily match the set values, such as humidity and temperature, can be obtained based on the measurement results of the corresponding measuring device 51 (e.g., a hygrometer and a thermometer). On the other hand, for other state variables whose actual values ​​basically match the set values ​​(e.g., the amount of supplied air suction and electrostatic adsorption voltage), the control unit 50 can obtain them from data stored in the memory unit as setting parameter information (control parameters).

[0044] Furthermore, the term "transportation state" here refers to a set of "transportation conditions" and the "transportation result" actually obtained when the workpieces are transported under those transport conditions. In this embodiment, the "transportation result" includes the supply speed, which indicates the number of workpieces transported per unit time by the transport device 2, and the uninspected rate, which indicates the percentage of workpieces that were not inspected due to transport defects among multiple workpieces (= "number of workpieces that were not inspected due to transport defects" / "total number of workpieces").

[0045] Furthermore, "failure to transport" as used here refers to a situation where the workpiece is in a non-aligned state (e.g., position and orientation) that prevents it from being properly inspected. For example, if the workpiece deviates from the desired transport path 5, or if the orientation of the workpiece on the transport path 5 deviates from the desired orientation, this may be considered a "failure to transport."

[0046] In this way, the control unit 50 acquires transport status information, which includes a plurality of interrelated control parameters and cumulative data regarding transport results.

[0047] If it is difficult to obtain the "current transport conditions," such as when the inspection system 100 is not operating at all (stopped), then for the initial period of operation of the inspection system 100 (for example, the first one to several trials (for example, around 2 to 9 times)), any transport conditions determined by the adjuster (user) may be used as the "current transport conditions."

[0048] Then, after acquiring the transport state as described above, the control unit 50 estimates the transport state under arbitrary transport conditions (S2), and determines the next transport conditions based on the estimated transport state (S3).

[0049] In this context, "estimating the transport state" means predicting the expected value and variance of the transport results (including supply speed and uninspected rate) for any controllable transport condition based on the transport state data acquired and recorded so far, using Gaussian process regression in this embodiment. For example, the control unit 50 can estimate multiple transport states from the acquired transport state information using Gaussian process regression, including combination data of a set of "multiple control parameters" selected from a search range and the "uninspected rate" associated with that set of multiple control parameters.

[0050] Furthermore, in this embodiment, "determining the next transport conditions" means finding the conditions under which the uninspected rate is likely to be minimized, given that the supply speed is equal to or greater than the set speed. For example, the control unit 50 can determine, among the acquired transport conditions and estimated transport conditions, multiple control parameters for the transport condition under which the uninspected rate is minimized, given that the supply speed is equal to or greater than the set speed, as multiple control parameters included in the parameter set.

[0051] More specifically, the control unit 50 determines the following transport condition "x" based on the following formula. next It is possible to determine "x" here. next " is a k-dimensional vector (where "k" is any natural number representing the number of control parameters). Also, in the following equation, "μ" r (x) is the expected value of the "uninspected rate" of the work, and "σ r (x) is the standard deviation (square root of variance) of the "uninspected rate" of the work, and "μ v (x) is the expected value of the "supply rate" of the workpiece, and "σ v (x) is the standard deviation (square root of variance) of the workpiece's "supply rate". Also, "v spec" is the specified speed related to the "feeding speed" of the workpiece, and is determined by the adjuster (user). In the following formula, "α" and "β" are arbitrary real numbers determined by the adjuster, and can be set within the range of "0" to "3".

number

[0052] As described above, the control unit 50 uses Gaussian regression to derive a parameter set containing multiple control parameters that exhibits the minimum uninspected rate from the transport state information. That is, the control unit 50 uses Gaussian regression to derive a parameter set containing multiple control parameters that exhibits the minimum uninspected rate under the condition that the workpiece supply speed is equal to or greater than the set speed. By using the workpiece supply speed as a condition in this way, it is possible to eliminate an inherently undesirable situation, such as "the uninspected rate approaching zero as a natural consequence of operating the device slowly by excessively slowing the workpiece supply speed."

[0053] After the following transport conditions (parameter set) are determined as described above, the control parameters of the various devices of the inspection system 100 (especially the various devices that adjust the state quantities affecting the transport of the workpiece (including the transport device 2 and peripheral devices)) are changed to the following transport conditions (parameter set) (S4). That is, the control parameters (transport conditions) of the various devices of the inspection system 100 are changed to the above "x next It will be changed to ".

[0054] The control parameters that are changed here may include control parameters that can be changed by the control unit 50 and control parameters that cannot be changed by the control unit 50 and must be changed by the adjuster (user). The control parameters that can be changed by the control unit 50 are updated by the control unit 50 to the "next transport conditions (parameter set)". On the other hand, for control parameters that are required to be changed by the adjuster (for example, the amount of gap between the work supply device 1 and the transport table 2a), the control unit 50 prompts the adjuster (user) to make the change via a notification device (for example, a user interface 52 such as a display or speaker). Therefore, the above processing step S4 may include "the control unit 50 starting to change the control parameters that can be changed in order to change to the next transport conditions" and "the control unit 50 prompting the user to change the control parameters that cannot be changed in order to change to the next transport conditions".

[0055] Furthermore, even for control parameters that can be changed by the control unit 50, such as humidity and temperature, it may take time for the change to actually be completed. Also, control parameters that are changed by the user will not be changed unless the user actually performs the change adjustment work. Thus, a time lag may occur between the timing when the control unit 50 performs the processing to change to the next transport conditions and the timing when the change to the next transport conditions is actually completed, and this time lag may become long enough to adversely affect transport control and inspection control.

[0056] Therefore, in the processing step S4, the control unit 50 may determine whether or not the change in transport conditions has been completed. In this case, if it is determined that the change in transport conditions has been completed, the series of steps described above (see S1 to S4) may be repeated.

[0057] On the other hand, if it is determined that the change in transport conditions has not been completed, the control unit 50 may determine whether or not an interruption instruction has been entered by the user. An "interruption instruction" here refers to a command entered by the user via the user interface 52, which is a command instructing the control unit 50 to interrupt the process of changing the transport conditions, even if the change in transport conditions has not been completed. The control unit 50 may then forcibly repeat the series of steps described above (see S1 to S4). Therefore, if it is determined that an interruption instruction has been entered, the control unit 50 may repeat the series of steps described above (see S1 to S4), even if it is determined that the change in transport conditions has not been completed.

[0058] On the other hand, if it is determined that no interruption instruction has been input, the control unit 50 may determine whether the elapsed time has exceeded the set elapsed time. Here, "elapsed time" refers to the time elapsed from the time when the process for "changing to the next transport condition" is started, and is measured by the timer device 61. If it is determined that the elapsed time has exceeded the set elapsed time, the control unit 50 may repeat the above series of steps (see S1 to S4) even if it is determined that the change in transport condition has not been completed. On the other hand, if it is determined that the elapsed time has not exceeded the set elapsed time, the control unit 50 may continue to perform the process of changing the control parameters that the control unit 50 can change in order to change to the next transport condition. The determination of whether the elapsed time has exceeded the set elapsed time may be made by the control unit 50 (timer device 61 in Figure 3), or by a timer device provided separately from the control unit 50.

[0059] If the change to the next transport condition cannot be completed immediately, the control unit 50 waits for the change to the next transport condition to be completed until an interruption instruction is input or until a certain time (set elapsed time) has elapsed, and during this waiting period, steps S1 to S4 described above do not need to be repeated again. If an interruption instruction is input or a certain time (set elapsed time) has elapsed, steps S1 to S4 described above may be forcibly repeated again. When steps S1 to S4 described above are forcibly repeated in this manner, the "current transport condition" in step S1 after the repetition does not necessarily match the "next transport condition" determined in step S3 before the repetition, and the transport condition at the time when step S1 after the repetition is executed is acquired as the "current transport condition".

[0060] In this way, based on the parameter set of multiple control parameters derived, multiple control parameters for each of the transport device 2 and peripheral equipment are updated, and the transport device 2 and peripheral equipment are operated based on the updated control parameters. The control unit 50 then acquires the latest multiple control parameters and the latest transport results, and derives a new parameter set from the transport status information, which includes cumulative data reflecting the latest multiple control parameters and the latest transport results. Based on the derived new parameter set, multiple control parameters for each of the transport device 2 and peripheral equipment are further updated, and the transport device 2 and peripheral equipment are operated based on the further updated control parameters.

[0061] According to the transport control flow described above, data on the transport status under the operation of various devices of the inspection system 100 is acquired, recorded, and stored. Based on the recorded and stored transport status data, transport status data that has not been recorded or stored is estimated. Based on the recorded and stored transport status data and the estimated transport status data, transport conditions that are considered to show the minimum uninspected rate (estimated ideal transport conditions) are estimated. Then, the various devices of the inspection system 100 are operated under the estimated ideal transport conditions, data on the transport status under the estimated ideal transport conditions is acquired, recorded, and stored. The estimation of transport status that has not been recorded or stored and the estimation of estimated ideal transport conditions are repeatedly performed.

[0062] By repeatedly performing this series of processes, it becomes possible to efficiently find the true ideal transport conditions, or transport conditions that are very close to the true ideal conditions, which result in the lowest uninspected rate, with minimal human intervention, which is advantageous in achieving more appropriate workpiece transport.

[0063] In particular, as the number of repetitions of the series of processes in the control flow described above increases, the data on the actual transport conditions and transport state that are acquired are cumulatively recorded and stored and used, making it possible to bring the transport state of the workpiece closer to the ideal transport state.

[0064] As described above, the transport control flow of this embodiment makes it possible to automate or nearly automate the search for optimal transport conditions, minimizing the need for the adjuster (user) to be constantly monitoring the equipment and manually adjusting control parameters. This also reduces the man-hours required for equipment adjustment and lessens the adjuster's workload. As a result, the decline in the production capacity of the inspection system 100 can be suppressed. Furthermore, the effort required for adjustment work when constructing the inspection system 100 can be reduced, which in turn can shorten the delivery time of the inspection system. In addition, since the adjustment of the transport state is performed efficiently based on Gaussian process regression, rather than relying on the adjuster's intuition and experience, variations in the adjustment accuracy of the control parameters of various devices can be suppressed.

[0065] Also, the transfer control flow of the present embodiment is also advantageous for maintaining optimal transfer conditions. That is, it is possible to suppress the occurrence of disturbances in the transfer posture of the work during the operation of various devices in the inspection system 100, substantially improve the operation rate of various devices, and as a result, reduce the un-inspected rate and increase productivity, thereby improving user satisfaction.

[0066] [Gaussian process regression] Next, an example of a method for estimating the transfer state of a work based on Gaussian process regression will be described.

[0067] According to Gaussian process regression, it is possible to perform estimation considering noise (degree of uncertainty of prediction), and it is also possible to estimate a non-linear function. Also, since Gaussian process regression is a method of machine learning, it is expected that the estimation accuracy will be improved by repeating trials and collecting and learning data.

[0068] In the present embodiment, the expected value and standard deviation of the un-inspected rate can be derived based on, for example, the following formula by Gaussian process regression. The following formula is based on the expected value (μ r (x)) and standard deviation (σ r (x)) of the "un-inspected rate", but the expected value (μ v (x)) and standard deviation (σ v (x)) of the "supply rate" of the work can also be derived based on a formula obtained by appropriately substituting "r (un-inspected rate)" in the following formula with "v (supply rate)".

Equation

[0069] The kernel function can be said to be a function that defines how similar x (m-dimensional vector) and x' (m-dimensional vector) are to each other. For example, a function of the following formula can be used as the kernel function.

[0070]

Equation

[0071] However, if we simply take the distance between the two inputs (x and x') (i.e., take the square root of the sum of the squared differences between the elements of the two inputs), we cannot compare the scales because the dimensions of each element of x (an m-dimensional vector) are different, and we cannot obtain a meaningful value. Therefore, we divide each element of x (an m-dimensional vector) by an appropriate constant for each element, and then calculate the distance. The constant used here is determined empirically by the adjuster, and a value approximately equal to the width of the domain of each element can be used as the constant.

[0072] Furthermore, "c" and "l" are constants that can be appropriately determined by the adjuster (user). For example, the recorded transport status (x0... x n The constants "c" and "l" may be chosen such that the likelihood generated from the assumed Gaussian distribution by the Gaussian process is maximized. However, since it is generally difficult to find an exact solution, "c" and "l" may be determined based on any numerical method such as Newton's method. Note that it is not necessary to perform the calculations to determine "c" and "l" each time; reasonable values ​​determined in advance based on appropriately prepared data can be used as "c" and "l".

[0073] [Example of Gaussian process regression simulation] The inventor of this case actually performed a simulation on a computer to determine a single transport result (1D output) called "uninspected rate (no_insp)" using four state variables (4D inputs) as transport conditions (control parameters): "electrostatic adsorption voltage (vol)", "supply air intake amount (air)", "temperature (tmp)", and "humidity (hum)". In this simulation, the "supply speed" was always set to the specified speed (v spec It was assumed that the condition of being ) or higher was met.

[0074] In other words, once the inputs (vol, air, tmp, hum) were determined, the simulation was performed by repeatedly executing an operation (referred to here as a "trial") that yielded a single output (no_insp) via an unknown function.

[0075] In this simulation, the objective was to find the input values ​​that yield the smallest possible uninspected rate, given that the formula (function) representing the uninspected rate was unknown, and using the smallest possible number of trials.

[0076] For the unknown function, the following formula was used in this simulation, but other formulas may also be used.

number

[0077] The minimum and maximum values ​​for each input value (vol, air, tmp, hum) within the possible range (search range) are shown in the table below. As is clear from the above formula, the optimal values ​​for each input value (vol, air, tmp, hum) that yield the minimum output value (no_insp; untested rate) are also listed in the table below. [Table 1]

[0078] The simulation, under the assumption that the optimal values ​​of the inputs (vol, air, tmp, hum) are unknown, repeatedly attempts to find each input value that is considered optimal within the search range determined by the minimum and maximum values ​​mentioned above. The evaluation was then based on how close the resulting estimated optimal input value was to the true optimal value mentioned above.

[0079] Specifically, since there is no information about the transport status at the start of the simulation, for the first P trials (where "P" is an arbitrary natural number (in this example, P=5); the initial trial process), values ​​randomly selected from within the search range were used as input values.

[0080] Subsequently, the results obtained in this initial trial process (information on the transport state (transport conditions and transport results)) were used to conduct trials based on Gaussian process regression (Gaussian trial process). Specifically, based on the above equation (see "Equation 1"), Q trials (where "Q" is an arbitrary natural number (in this example, Q=30)) were repeated to find the input values ​​(vol, air, tmp, hum) that minimize "μ-3σ".

[0081] In other words, the mean μ and standard deviation σ of the output (untested rate) for each input were determined. The objective is to find input values ​​within the search range that result in a small mean μ, but areas with a large standard deviation σ are regions where the prediction is not sufficiently accurate. Therefore, based on the above formula (see "Equation 1"), the input value that minimizes "μ - mσ (m=3 in this example)" was selected as the search point for the next trial. After the trial is performed for the search point determined in this way, Gaussian process regression is performed again using the trial results of the search point obtained so far to determine the next search point. In the Gaussian trial process, this series of processes is repeated, and as the number of trials increases (i.e., as the search progresses), the region with a large standard deviation σ is gradually reduced, and the prediction accuracy improves.

[0082] Subsequently, the input values ​​(vol, air, tmp, hum) associated with the minimum output value (no_insp) predicted from the trial results of the search points in the initial trial process and the Gaussian trial process described above were taken as the optimal input values ​​obtained from this simulation. The table below shows the optimal input values ​​obtained from this simulation ("predicted optimal values") and the "true optimal values" determined based on the above function formula (see "Equation 4"). [Table 2]

[0083] As is clear from the results above, the Gaussian process regression of this embodiment makes it possible to obtain a predicted optimal value (control parameter) that is very close to the true optimal value.

[0084] In particular, among the four types of inputs mentioned above, the electrostatic adsorption voltage (vol) and the supplied air intake amount (air) are control parameters that can be adjusted by the control unit 50 without the intervention of the operator (user). Therefore, for the electrostatic adsorption voltage (vol) and the supplied air intake amount (air), the control unit 50 can automatically search for optimal control parameters and adjust them to optimal parameters. Furthermore, for temperature (tmp) and humidity (hum), the derivation of control parameters by Gaussian process regression in this embodiment can be useful in determining whether they are input elements that do not have a significant impact on the uninspected rate due to transport defects or whether they have a significant impact.

[0085] As described above, this embodiment makes it possible to optimize all control conditions (input values; control parameters) at once. Compared to conventional methods in which the adjuster (user) adjusts multiple control conditions one by one, this significantly reduces the burden on the user and is also advantageous in terms of stabilizing adjustment accuracy.

[0086] In particular, with conventional technology, an increase in the number of adjustment steps reduces the production capacity of the inspection system. Furthermore, the personalization of adjustment work leads to variations in the quality of adjustment (and consequently the quality of inspection) and difficulty in securing personnel, which can further reduce the production capacity of the inspection system. Moreover, operating the inspection system while waiting for an adjustment worker (i.e., when the inspection system is not in good condition) can also lead to a decrease in the production capacity of the inspection system. On the other hand, the inspection system 100 of this embodiment is effective in avoiding or suppressing an increase in the number of adjustment steps and personalization of adjustment work, and can prevent a decrease in the production capacity of the inspection system.

[0087] As described above, the inspection system 100 of this embodiment includes a transport device 2 that transports a plurality of workpieces W, an inspection device that inspects the plurality of workpieces W being transported by the transport device 2, peripheral equipment that adjusts state variables affecting the transport of the plurality of workpieces W, and a control unit 50 that acquires transport results including a supply speed indicating the number of workpieces W transported per unit time by the transport device 2 and an uninspected rate indicating the percentage of workpieces W that were not inspected due to transport defects among the plurality of workpieces W, and the control unit 50 controls the operation of the transport device 2 and peripheral equipment by adjusting a plurality of control parameters for each of the transport device 2 and peripheral equipment. The control unit 50 acquires transport state information including a plurality of interrelated control parameters and cumulative data regarding the transport results, derives a parameter set that includes a plurality of control parameters and shows the smallest uninspected rate from the transport state information using Gaussian process regression, and updates a plurality of control parameters for each of the transport device 2 and peripheral equipment based on the derived parameter set.

[0088] Furthermore, the inspection method of this embodiment includes the steps of transporting a plurality of workpieces W by a transport device 2, inspecting the plurality of workpieces W being transported by the transport device 2 using an inspection device, and obtaining transport results including a supply speed indicating the number of workpieces W transported per unit time by the transport device 2 and an uninspected rate indicating the percentage of workpieces W that were not inspected due to transport defects among the plurality of workpieces W, and adjusting a plurality of control parameters for each of the peripheral equipment and the transport device 2 to adjust state quantities that affect the transport of the plurality of workpieces W, thereby controlling the operation of the transport device 2 and peripheral equipment. In the step of adjusting a plurality of control parameters to control the operation of the transport device 2 and peripheral equipment, transport state information including a plurality of interrelated control parameters and cumulative data regarding the transport results is obtained, a parameter set including a plurality of control parameters that shows the smallest uninspected rate is derived from the transport state information using Gaussian process regression, and a plurality of control parameters for each of the transport device 2 and peripheral equipment are updated based on the derived parameter set.

[0089] Furthermore, the inspection method of this embodiment may be implemented as a program. Such a program is a program that causes a computer to execute a procedure for transporting a plurality of workpieces W by the transport device 2, a procedure for inspecting the plurality of workpieces W being transported by the transport device 2 by the inspection device, a procedure for obtaining transport results including a supply speed indicating the number of workpieces W transported per unit time by the transport device 2 and an uninspected rate indicating the percentage of workpieces W that were not inspected due to transport defects, and a procedure for adjusting a plurality of control parameters for the peripheral device and the transport device 2, respectively, to adjust state quantities that affect the transport of the plurality of workpieces W, thereby controlling the operation of the transport device 2 and the peripheral device. In the procedure for adjusting a plurality of control parameters to control the operation of the transport device 2 and the peripheral device, transport state information including a plurality of interrelated control parameters and cumulative data regarding the transport results is obtained, a parameter set including a plurality of control parameters that shows the smallest uninspected rate is derived from the transport state information using Gaussian process regression, and a plurality of control parameters for the transport device 2 and the peripheral device, respectively, are updated based on the derived parameter set.

[0090] Furthermore, the peripheral equipment may include a first peripheral device having a charging section 6 that electrostatically attracts the workpiece W to the transport device 2 with an attractive force corresponding to the applied voltage, and a second peripheral device having vacuum suction sections 17 and 18 that suck up the workpiece W with an attractive force corresponding to the air flow rate. The parameter set may also include control parameters related to the applied voltage to the charging section 6 and control parameters related to the air flow rate in the vacuum suction sections 17 and 18.

[0091] Furthermore, the control unit 50 may use Gaussian process regression to estimate multiple transport states from the transport state information, including combination data of multiple control parameters selected from the search range of multiple control parameters and the uninspected rate associated with the multiple control parameters. Among the multiple estimated transport states, the control parameters of the transport state in which the uninspected rate is minimized under the condition that the supply speed is equal to or greater than the set speed may be used as multiple control parameters included in the parameter set.

[0092] The control unit 50 may also acquire the latest multiple control parameters and the latest transport results after the update, and may derive a parameter set from transport status information including cumulative data that reflects the latest multiple control parameters and the latest transport results.

[0093] Furthermore, the state quantities that affect the transport of multiple workpieces W may include one or more of the following: the rotational speed of the transport table on which the transport device 2 has and on which the multiple workpieces W are placed; the airflow rate in the vacuum suction units 17 and 18 that suck each of the multiple workpieces W when they are placed on the transport table; the voltage applied to the charging unit 6 that electrostatically attracts each of the multiple workpieces W to the transport table; the humidity around the transport table; the temperature around the transport table; the drive parameters of the parts supply device that supplies the multiple workpieces W to the transport table; the amount of gap between the parts supply device and the transport table; and the type of alignment guide that aligns each of the multiple workpieces W on the transport table.

[0094] It should be noted that the embodiments and modifications disclosed herein are illustrative in all respects and should not be construed restrictively. The embodiments and modifications described above may be omitted, substituted, and modified in various ways without departing from the scope and spirit of the appended claims. For example, the embodiments and modifications described above may be combined in whole or in part, and other embodiments may be combined with the embodiments or modifications described above. Furthermore, the effects described herein are illustrative, and other effects may result.

[0095] The technical categories that embody the above-described technical concept are not limited. For example, the above-described technical concept may be embodied by a computer program that causes a computer to execute one or more steps included in a method for manufacturing or using the above-described device. Alternatively, the above-described technical concept may be embodied by a computer-readable, non-transitory recording medium on which such a computer program is recorded. [Explanation of symbols]

[0096] 1 Work supply device, 1a Linear feeder, 2 Conveying device, 2a Conveying table, 2b Conveying drive device, 2s Top surface, 4 Vibration-free section, 5 Conveying path, 6 Charged section, 7 Alignment guide, 7a Guide surface, 7z Bottom surface, 8 Side camera section, 9 Inner camera section, 10 Top camera section, 11 Bottom camera section, 12 Front camera section, 13 Rear camera section, 14 Discharge section, 17 Suction device, 17a Relay line, 18 Suction passage, 18a One end, 18b Other end, 20 Imaging section, 21 Transfer and alignment section, 50 Control section, 51 Measurement device, 52 User interface, 55 Various equipment, 56 External storage section, 60 Image analysis section, 61 Timer device, 62 Internal storage section, 100 Inspection system, W Work, γ Gap

Claims

1. A transport device for transporting multiple electronic components, An inspection device for inspecting the plurality of electronic components being transported by the transport device, A peripheral device that adjusts state variables affecting the transport of the aforementioned multiple electronic components, A control unit that acquires a transport result including a supply rate indicating the number of electronic components transported per unit time by the transport device and an uninspected rate indicating the percentage of electronic components among the plurality of electronic components that were not inspected due to transport failure, comprising a control unit that controls the operation of the transport device and the peripheral device by adjusting a plurality of control parameters relating to the transport device and the peripheral device, respectively. The control unit, Acquire transport status information including the multiple control parameters that are related to each other and cumulative data regarding the transport results, Using Gaussian process regression from the transport status information, a parameter set containing the plurality of control parameters that exhibits the smallest rate of uninspected items is derived. Based on the derived parameter set, the multiple control parameters for the transport device and the peripheral equipment are updated. Inspection system.

2. The aforementioned peripheral devices are A first peripheral device having a charging section that electrostatically attracts electronic components to the transport device with an adsorption force corresponding to the applied voltage, The device includes a second peripheral device having a vacuum suction section that sucks up electronic components with a suction force corresponding to the airflow rate, The parameter set includes control parameters relating to the applied voltage to the charging section and control parameters relating to the airflow rate in the vacuum suction section. The inspection system according to claim 1.

3. The control unit, From the transport state information, Gaussian process regression is used to estimate multiple transport states, which include combination data of the multiple control parameters selected from the search range of the multiple control parameters and the uninspected rate associated with the multiple control parameters. Among the estimated multiple transport conditions, the control parameters of the transport condition in which the uninspected rate is minimized under the condition that the supply speed is equal to or greater than the set speed are used as the multiple control parameters included in the parameter set. The inspection system according to claim 1.

4. The control unit, The latest updated control parameters and the latest transport results are obtained. A parameter set is derived from transport status information, which includes the latest control parameters and cumulative data reflecting the latest transport results. The inspection system according to claim 1.

5. The state quantities affecting the transport of the plurality of electronic components include one or more of the following: the rotational speed of the transport table on which the transport device has and on which the plurality of electronic components are placed; the airflow rate in the vacuum suction section that sucks each of the plurality of electronic components when they are placed on the transport table; the voltage applied to the charging section that electrostatically attracts each of the plurality of electronic components to the transport table; the humidity around the transport table; the temperature around the transport table; the drive parameters of the component supply device that supplies the plurality of electronic components to the transport table; the amount of clearance between the component supply device and the transport table; and the type of alignment guide that aligns each of the plurality of electronic components on the transport table. The inspection system according to claim 1.

6. A process of transporting multiple electronic components using a transport device, A step of inspecting the plurality of electronic components being transported by the transport device using an inspection device, The process includes obtaining a transport result that includes a supply speed indicating the number of electronic components transported per unit time by the transport device and an uninspected rate indicating the percentage of electronic components among the plurality of electronic components that were not inspected due to transport defects, and then adjusting a plurality of control parameters related to the peripheral device and the transport device, respectively, which adjust the state quantities that affect the transport of the plurality of electronic components, in order to control the operation of the transport device and the peripheral device. In the step of adjusting the plurality of control parameters to control the operation of the transport device and the peripheral equipment, Transport status information is acquired, which includes the multiple control parameters that are related to each other and cumulative data regarding the transport results. From the transport status information, a parameter set including the plurality of control parameters that shows the smallest uninspected rate is derived using Gaussian process regression. Based on the derived parameter set, the multiple control parameters for the transport device and the peripheral equipment are updated. Testing method.

7. On the computer, Procedures for transporting multiple electronic components using a transport device, A procedure for inspecting the plurality of electronic components being transported by the transport device using an inspection device, A program for executing a procedure to control the operation of a transport device and a peripheral device by adjusting a plurality of control parameters relating to a peripheral device and the transport device, respectively, which adjust state quantities that affect the transport of the plurality of electronic components, by obtaining transport results including a supply speed indicating the number of electronic components transported per unit time by the transport device and an uninspected rate indicating the percentage of electronic components among the plurality of electronic components that were not inspected due to transport failure, and adjusting a plurality of control parameters relating to a peripheral device and the transport device, respectively, which adjust the state quantities that affect the transport of the plurality of electronic components, In the procedure for controlling the operation of the transport device and peripheral equipment by adjusting the aforementioned plurality of control parameters, Transport status information is acquired, which includes the multiple control parameters that are related to each other and cumulative data regarding the transport results. From the transport status information, a parameter set including the plurality of control parameters that shows the smallest uninspected rate is derived using Gaussian process regression. Based on the derived parameter set, the multiple control parameters for the transport device and the peripheral equipment are updated. program.

Images