Integrated AOI and machine learning for SMT pick and place machines

Abstract

A pick and place machine includes a gantry system, a pick and place head movable by the gantry system, an automated optical inspection (AOI) camera mounted within the pick and place machine and configured to capture images of an entire printed circuit board assembly (PCBA), and a processor. The processor is configured to analyze the images to detect placement defects and calculate accuracy offsets for components placed on the PCBA, and modify a placement recipe of the pick and place machine based on the calculated accuracy offsets to improve placement accuracy.

Description

INTEGRATED AOI AND MACHINE LEARNING FOR SMT PICK AND PLACE MACHINESRELATED APPLICATION

[0001] The present application claims priority to U.S. Provisional Patent Application No. 63 / 687,967 filed on August 28, 2024 and titled “LOCAL AOI AND MACHINE LEARNING WITHIN AN SMT PICK AND PLACE MACHINE” the entirety of which is incorporated by reference herein.FIELD OF INVENTION

[0002] The present disclosure relates to surface mount technology (SMT) manufacturing systems, and more particularly to SMT pick and place machines incorporating integrated automated optical inspection (AOI) and machine learning capabilities for improved component placement accuracy.BACKGROUND

[0003] Surface mount technology (SMT) has become a cornerstone of modem electronics manufacturing, enabling the production of compact and highly integrated electronic devices. SMT pick and place machines are integral to this process, responsible for accurately positioning and placing electronic components onto printed circuit boards (PCBs). As electronic devices continue to shrink in size while increasing in complexity', the demands on SMT pick and place machines have grown correspondingly.

[0004] The accuracy of component placement is a key factor in the quality and reliability of finished electronic products. Even minor deviations in placement can lead to defects, reduced performance, or complete failure of the assembled device. Consequently, manufacturers are constantly seeking ways to improve the precision and consistency of SMT pick and place operations.

[0005] SMT pick and place machines are typically assembled into a line of multiple and different machines. It is common that these lines contain an AOI (Automated Optical Inspection) machine as a standalone piece of equipment in the line. AOI machines are typically found at the end of an SMT line, post-reflow, and less frequently found pre-reflow as well. Within pick and place machines themselves there is typically one or more cameras mounted to the positioning system(s) which is used for several functions. The primary function of these cameras is pattern error correction (PEC) which images a PCBA's (printed circuit board assembly) fiducials (or other reference marks) to locate the PCBA in machine space. At a minimum, this PEC inspection requires the cameras to inspect two fiducials, but this quantity of fiducials is typically greaterthan three and can go up to tens of fiducials. For every fiducial found the gantry must position the camera over the fiducial to then acquire and process and image.

[0006] This then is used to adjust the placement pattern to enable accurate placements. These cameras do have the capability of performing local inspections of component placements. These local inspections of component placements are done one component placement at a time in a process called post bond inspect or post bond inspection.

[0007] Using prior art AOI machine(s), when using an AOI machine post reflow, it can be difficult to determine whether there is a placement issue caused by the pick and place machine or some other root cause such as reflow, solder printing, etc. In such systems, each AOI machine is an independent machine which increases overall line length and consumes valuable factory floorspace. Moreover, each AOI machine is a large added expense to a line of equipment. Each AOI machine requires higher cost of ownership and operation (more service, more operators, etc.). Because AOI equipment suppliers frequently do not also provide SMT equipment (and vice versa), collating of data can be difficult.

[0008] Using PEC cameras and post bond inspections, existing methodologies have a large impact on cycle time because the inspection is performed one component at a time and performed in series with the placement routines. For instance, the machine cannot pick while the images are being taken because the camera used for inspection is affixed directly to the placement machine.

[0009] Moreover, machine learning and artificial intelligence technologies have shown promise in various manufacturing applications, offering the potential to analyze complex data patterns and make real-time adjustments to improve processes. In the context of SMT assembly, these technologies could potentially be leveraged to enhance placement accuracy and defect detection.

[0010] Despite ongoing advancements, challenges remain in achieving consistent, high- precision component placement across large production volumes. Factors such as thermal expansion, mechanical wear, and variations in component tolerances can all contribute to placement errors over time. Additionally, the increasing miniaturization of electronic components places ever-greater demands on the precision capabilities of SMT equipment.

[0011] As the electronics industry continues to evolve, there is an ongoing need for innovations that can improve the accuracy, efficiency, and reliability of SMT pick and place operations. Addressing these challenges could lead to improvements in product quality, reduced manufacturing costs, and the ability to produce increasingly sophisticated electronic devices.SUMMARY

[0012] In one aspect, a pick and place machine includes: a gantry' system; a pick and place head movable by the gantry' system; an automated optical inspection (AOI) camera mounted within the pick and place machine and configured to capture images of an entire printed circuit board assembly (PCBA); and a processor configured to: analyze the images to detect placement defects and calculate accuracy offsets for components placed on the PCBA, and modify a placement recipe of the pick and place machine based on the calculated accuracy offsets to improve placement accuracy.

[0013] Additionally or alternatively, the AOI camera is configured to capture a baseline image of the PCBA before component placement and a second image after component placement.

[0014] Additionally or alternatively, the processor is further configured to: compare the baseline image and the second image to determine locations of components placed relative to reference points on the PCBA.

[0015] Additionally or alternatively, the processor is further configured to: present a message to a user to verify detected placement defects as true failures or false failures.

[0016] Additionally or alternatively, the processor is further configured to: perform machine learning to improve defect detection capabilities based on user verification of detected placement defects.

[0017] Additionally or alternatively, the processor is further configured to: compile and analyze placement accuracy data from multiple PCBAs to identify statistical trends in placement errors.

[0018] Additionally or alternatively, the processor is further configured to: perform root cause analysis of placement errors based on the compiled placement accuracy data and machine characteristic data.

[0019] In another aspect, a method of operating a pick and place machine includes: capturing, using an automated optical inspection (AOI) camera mounted within the pick and place machine, a baseline image of a printed circuit board assembly (PCBA) to determine locations of pad sites relative to reference marks; placing components on the PCBA; capturing, using the AOI camera, a post-placement image of the PCBA; analyzing the post-placement image to detect placement defects and calculate accuracy offsets for the placed components; and modifying a placement recipe of the pick and place machine based on the calculated accuracy offsets.

[0020] Additionally or alternatively, the method further includes: presenting a message to a user to verify detected placement defects as true failures or false failures.

[0021] Additionally or alternatively, the method further includes: performing machine learning to improve defect detection capabilities based on user verification of detected placement defects.

[0022] Additionally or alternatively, the method further includes: compiling and analyzing placement accuracy data from multiple PCBAs to identify statistical trends in placement errors.

[0023] Additionally or alternatively, the method further includes: performing root cause analysis of placement errors based on the compiled placement accuracy data and machine characteristic data.

[0024] Additionally or alternatively, analyzing the post-placement image includes: comparing the baseline image and the post-placement image to determine locations of components placed relative to the reference marks on the PCBA.

[0025] Additionally or alternatively, modify ing the placement recipe includes: adjusting component placement coordinates based on the determined locations of components relative to the reference marks.

[0026] In another aspect, a system for improving placement accuracy in printed circuit board assembly manufacturing includes: a plurality of pick and place machines arranged in a production line; an automated optical inspection (AOI) camera mounted within at least one of the pick and place machines; and a processor configured to: analyze images captured by the AOI camera to detect placement defects and calculate accuracy offsets for components placed on printed circuit board assemblies (PCBAs), and distribute placement recipe modifications based on the calculated accuracy offsets to the plurality of pick and place machines.

[0027] Additionally or alternatively, the AOI camera is configured to capture a baseline image of each PCBA before component placement and a post-placement image after component placement.

[0028] Additionally or alternatively, the processor is further configured to: compare the baseline image and the post-placement image to determine locations of components placed relative to reference marks on each PCBA.

[0029] Additionally or alternatively, the processor is further configured to: adjust component placement coordinates in the placement recipe modifications based on the determined locations of components relative to the reference marks.

[0030] Additionally or alternatively, the processor is further configured to: compile and analyze placement accuracy data from multiple PCBAs across the plurality of pick and place machines to identify statistical trends in placement errors.

[0031] Additionally or alternatively, the processor is further configured to: perform root cause analysis of placement errors based on the compiled placement accuracy data and machine characteristic data from each of the plurality of pick and place machines.BRIEF DESCRIPTION OF FIGURES

[0032] The above and further advantages of this invention may be better understood by referring to the following description in conjunction with the accompanying drawings, in which like numerals indicate like structural elements and features in the various figures. For clarity, not every element may be labeled in every figure. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention.

[0033] FIG. 1 depicts a block diagram of an SMT Pick and Place Machine, in accordance with one embodiment.

[0034] FIG. 2 depicts a block diagram of an SMT production line, in accordance with one embodiment.

[0035] FIG. 3A depicts a top view of a pick and place machine, in accordance with one embodiment.

[0036] FIG. 3B depicts a side schematic view of the surface mount technology pick and place machine of FIG. 3 A, in accordance with one embodiment.

[0037] FIG. 4 depicts a block diagram of an AOI system, in accordance with one embodiment.

[0038] FIG. 5 depicts a block diagram of a surface mount technology’ (SMT) production line system, in accordance with one embodiment.

[0039] FIG. 6 depicts a flowchart of a process for component placement and inspection in a pick and place machine, in accordance with one embodiment.DETAILED DESCRIPTION

[0040] Reference in the specification to an embodiment or example means that a particular feature, structure or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the teaching. References to a particular embodiment or example within the specification do not necessarily all refer to the same embodiment or example.

[0041] The present teaching will now be described in detail with reference to exemplary embodiments or examples thereof as show n in the accompanying drawings. While the present teaching is described in conjunction with various embodiments and examples, it is not intended that the present teaching be limited to such embodiments and examples. On the contrary, the present teaching encompasses various alternatives, modifications, and equivalents, as will beappreciated by those of skill in the art. Moreover, features illustrated or described for one embodiment or example may be combined with features for one or more other embodiments or examples. Those of ordinary skill having access to the teaching herein will recognize additional implementations, modifications, and embodiments, as well as other fields of use, which are within the scope of the present disclosure as described herein.

[0042] In brief overview, the present disclosure provides systems and methods for integrating automated optical inspection (AOI) and machine learning capabilities within surface mount technology (SMT) pick and place machines. The disclosed systems and methods aim to improve component placement accuracy and defect detection in SMT manufacturing processes.

[0043] The integration of AOI cameras directly within pick and place machines allows for high-resolution imaging of printed circuit board assemblies (PCBAs) before and after component placement. This enables real-time monitoring and analysis of placement accuracy without the need for separate, standalone AOI equipment.

[0044] By capturing baseline images of PCBAs prior to component placement, the system establishes reference points for subsequent analysis. After components are placed, additional images are captured and compared to the baseline to detect placement defects and calculate accuracy offsets for each component.

[0045] Machine learning algorithms analyze the captured image data to identify placement errors, misalignments, and other defects. The system may present detected defects to operators for verification, using this feedback to continuously improve its defect detection capabilities over time.

[0046] Calculated accuracy offsets and identified placement errors are used to modify placement recipes in real-time, allowing the pick and place machines to dynamically adjust their operations for improved accuracy. This closed-loop feedback system enables continuous optimization of the placement process.

[0047] The integration of these capabilities within the pick and place machines themselves offers several advantages over traditional separate AOI systems. It reduces overall equipment footprint, lowers capital expenses, and enables faster detection and correction of placement issues. The system may compile and analyze placement data across multiple PCBAs and machines to identify statistical trends and perform root cause analysis of recurring errors.

[0048] By leveraging machine learning and integrated optical inspection, the disclosed systems and methods aim to enhance the efficiency, accuracy, and yield of SMT manufacturing processes. The continuous monitoring and optimization capabilities may lead to reduced defect rates, improved product quality, and increased overall manufacturing productivity.

[0049] In particular, by using the AOI systems described herein, pattern error correction may be performed, eliminating the need for up to two additional cameras on the pick and place machine and further reduces the time required for inspection. In such an embodiment, it is contemplated that one global image can be taken by the AOI camera and used for the purpose of PEC. Such an embodiment will also increase speed of the PEC process.

[0050] FIG. 1 illustrates a block diagram of an SMT Pick and Place Machine 100, in accordance with one embodiment. The SMT Pick and Place Machine 100 includes a Gantry System 102 containing a Pick and Place Head 104, an AOI Camera 106, and a Computer System 108 containing a Processor 110 and Data Storage 112. The system is configured to interact with a PCBA 114.

[0051] The Gantry System 102 may provide a framework for precise movement and positioning within the SMT Pick and Place Machine 100. In some cases, the Gantry System 102 may include linear motion components such as rails, bearings, and actuators to enable controlled movement in multiple axes.

[0052] A Pick and Place Head 104 may be movable by the Gantry System 102. The Pick and Place Head 104 may be responsible for picking up electronic components from a component supply area and placing them onto specific locations on a PCBA 114. In some cases, the Pick and Place Head 104 may include one or more nozzles for component handling and a vision system for component alignment.

[0053] An AOI Camera 106 may be mounted within the SMT Pick and Place Machine 100. The AOI Camera 106 may be configured to capture images of an entire PCBA 114. In some cases, the AOI Camera 106 may capture an image of the entire PCBA 114 when the positioning system of the pick and place machine is not obstructing the view. This configuration allows for comprehensive inspection of the PCBA 114 at various stages of the assembly process.

[0054] The Computer System 108 may include a Processor 110 and Data Storage 112. The Processor 110 may be configured to analyze images captured by the AOI Camera 106 to detect placement defects and calculate accuracy offsets for components placed on the PCBA 114. In some cases, the Processor 110 may employ image processing algorithms and machine learning techniques to perform these analyses.

[0055] Based on the calculated accuracy offsets, the Processor 110 may be configured to modify a placement recipe of the SMT Pick and Place Machine 100. This modification may involve adjusting component placement coordinates or other parameters to improve placement accuracy in subsequent operations.

[0056] The Data Storage 112 may store various types of data related to the operation of the SMT Pick and Place Machine 100. This may include placement recipes, image data, analysis results, and historical performance data.

[0057] By integrating the AOI Camera 106 directly within the SMT Pick and Place Machine 100, the system enables real-time monitoring and adjustment of component placement processes. This integration may lead to improved placement accuracy, reduced defect rates, and enhanced overall efficiency of the SMT assembly process.

[0058] FIG. 2 illustrates a block diagram of an SMT production line 200, in accordance with one embodiment. The SMT production line 200 may include multiple Pick and Place Machines, specifically a first Pick and Place Machine 202, a second Pick and Place Machine 208, and any additional number of Pick and Place Machines up through a last Pick and Place Machine 210. These Pick and Place Machines 202. 208, 210 may be arranged in sequence to perform component placement operations on PCBAs as they move through the production line.

[0059] The last Pick and Place Machine 210 may include an AOI Camera 204 and a Local Processor 206 or computer system. In some cases, the AOI Camera 204 may be mounted within the first Pick and Place Machine 202, enabling high-resolution imaging of PCBAs during the assembly process. The Local Processor 206 or computer system may process image data from the AOI Camera 204 and communicate with other components of the SMT production line 200. While not shown, any of the Pick and Place Machines 202, 208, 210 in the SMT production line 200 may also include an AOI camera, such as AOI Camera 204, and a local processor, such as Local Processor 206. Typically, an AOI camera may be placed on the final machine in a production line so that the placement from prior machines may be measured and then corrections can be determined and distributed back to the previous machines in the line. The choice of configuration may depend on specific production requirements, balancing comprehensive monitoring with system complexity and cost considerations.

[0060] A Computer System 212 may be connected to the Pick and Place Machines 202, 208, 210 in the SMT production line 200. The Computer System 212 may include several modules designed to analyze data, optimize processes, and improve overall production efficiency. In some embodiments, the modules 214, 216, 218, 220, 222, 224 shown in the Computer System 212 may be conducted in a central or separate computer system that is separate from the individual Pick and Place Machines 202, 208, 210 in the system or line. In other embodiments, the modules 214, 216, 218, 220, 222, 224 shown in the Computer System 212 may be performed or located locally within the individual Pick and Place Machines 202, 208, 210. Whatever the embodiment.the modules 214, 216, 218, 220, 222, 224 shown in the Computer System 212 may be configured to perform the AOI functions described herein.[00611 The Central Control System 212 and / or the computer system of the Pick and PlaceMachines 202, 208, 210 may include a Data Analysis Module 214. In some cases, the Data Analysis Module 214 may be configured to analyze images captured by the AOI Camera 204 to detect placement defects and calculate accuracy offsets for components placed on PCBAs. This analysis may involve advanced image processing techniques and comparison of pre- and postplacement images.

[0062] A Recipe Modification Module 216 may also be part of the Central Control System 212 and / or the computer system of the Pick and Place Machines 202, 208, 210. The Recipe Modification Module 216 may use the data from the Data Analysis Module 214 to generate placement recipe modifications. These modifications may be distributed to the plurality of SMT pick and place machines in the production line, allowing for dynamic adjustment of placement parameters based on real-time analysis.

[0063] The Central Control System 212 and / or the computer system of the Pick and Place Machines 202, 208, 210 may also include a baseline analysis module 218. In some cases, the baseline analysis module 218 may analyze, receive and / or process baseline PCBA conditions, establishing reference points for subsequent comparisons.

[0064] A message module 220 may be incorporated into the Central Control System 212 and / or the computer system of the Pick and Place Machines 202, 208, 210. The message module 220 may generate notifications or alerts based on the analysis results, potentially flagging issues for operator attention or verification.

[0065] The Central Control System 212 and / or the computer system of the Pick and Place Machines 202, 208, 210 may also feature a machine learning module 222. The machine learning module 222 may continuously improve defect detection and placement optimization algorithms based on accumulated data and feedback from the production process, as described herein.

[0066] A trend analysis module 224 may be included in the Central Control System 212 and / or the computer system of the Pick and Place Machines 202, 208, 210. The trend analysis module 224 may analyze data across multiple PCBAs and pick and place machines to identity statistical trends in placement errors or other production issues.

[0067] By integrating AOI capabilities directly within the Pick and Place Machines and centralizing data analysis and process optimization, the SMT production line 200 may achieve improved placement accuracy, reduced defect rates, and enhanced overall production efficiency.

[0068] FIG. 3 A illustrates a top view of a pick and place machine 300, in accordance with one embodiment. FIG. 3B illustrates a side schematic view of the pick and place machine 300, in accordance with one embodiment.

[0069] The pick and place machine 300 may include a gantry system 310 positioned above a PCBA zone 340. A pick and place head 320 may be mounted on the gantry system 310 for performing component placement operations. The pick and place machine 300 may comprise a fixed rail 330 and a movable rail 332 that define the boundaries of the placement area.

[0070] An AOI camera 350 may be mounted in a fixed position above the PCBA zone 340. The AOI camera 350 may be configured to capture images of a printed circuit board 360 positioned within the PCBA zone 340. In some cases, the AOI camera 350 may be positioned to capture images when the gantry system 310 and pick and place head 320 are moved away from the PCBA zone 340. The gantry system 310 may carry the pick and place head 320. The arrangement of the AOI camera 350 may allow for capturing images of the printed circuit board 360 both before and after component placement operations.

[0071] The fixed rail 330 and movable rail 332 may allow controlled movement of the gantry system 310. the pick and place head 320 and / or the board handling and automated board movement during component placement operations on the printed circuit board 360. In some cases, the fixed rail 330 and movable rail 332 may provide for this board handling movement and support the printed circuit board 360 within the PCBA zone 340.

[0072] In some cases, the pick and place machine 300 may utilize a 3D optical imaging station instead of a 2D imaging system. The 3D optical imaging station may be used to map the topography of the printed circuit board 360. Mapping of board topography may be used to optimize placement sequences and ensure there are no obstructions with the pick and place head 320 during component placement operations.

[0073] The fixed rails 330 and movable rail 332 are exemplary and represent any conveyance system that transfers a PCB into the machine and positions the PCB beneath the camera. Such a conveyance system may include a clamp which locks the PCB into a fixed position. The AOI cameras described herein may be mounted in a fixed location within the machine (separate from the gantries).

[0074] FIG. 4 illustrates a block diagram of an AOI system 400, in accordance with one embodiment. The AOI system 400 may include a pick and place machine 401 that defines a PCBA zone 410 where printed circuit board assemblies may be positioned for component placement operations.

[0075] An AOI camera 460 may be positioned to capture images of printed circuit board assemblies within the PCBA zone 410. In some cases, the AOI camera 460 may capture a baseline image of a PCBA to determine locations of pad sites relative to reference marks before component placement. The AOI camera 460 may then capture a second image after components have been placed on the PCBA.

[0076] The AOI camera 460 may capture the image of the PCBA zone 410, which is then processed by a computer system (as described in FIG. 2) in order to output a quality offset determination 470. In some cases, the quality offset determination 470 may compare the baseline image and the second image to determine locations of components placed relative to reference points on the PCBA.

[0077] The quality offset determination 470 may then be fed back to the computer system as offset data feedback 480. which may contain placement parameters and operating instructions for the pick and place machine 401. The offset data feedback 480 may provide updated placement coordinates and parameters back to the pick and place machine 401 based on the analysis perfonned to generate the quality offset determination 470.

[0078] In some cases, a placement recipe of the pick and place machine 401 may be modified based on the calculated accuracy offsets provided by the offset data feedback 480. This modification may improve placement accuracy in subsequent operations.

[0079] The AOI system 400 may also be used for additional purposes. In some cases, the system may be used for barcode reading of the PCBA. The system may also be used to preinspect fiducial to pad locations for new product introductions to teach pad placements without the need of PCBA CAD data.

[0080] The arrangement of the AOI system 400 may allow for continuous monitoring of placement operations through the AOI camera 460, analysis of placement accuracy through the quality offset determination 470. and adjustment of placement parameters through the offset data feedback 480 to optimize the performance of the pick and place machine 401.

[0081] FIG. 5 illustrates a block diagram of an AOI system 500 for multiple pick and place machines, in accordance with one embodiment. The AOI system 500 may include a first pick and place machine 501, a second pick and place machine 502, and any number of additional pick and place machines 503 arranged in sequence along a board flow through line 550.

[0082] A first PCBA zone 510 may be associated with the first pick and place machine 501, a second PCBA zone 511 may be associated with the second pick and place machine 502, and a third PCBA zone 512 may be associated with the third pick and place machine 503. The board flow through line 550 may facilitate the movement of PCBAs through each PCBA zone.

[0083] A first AOI camera 560 may be positioned to capture images of PCBAs within the first PCBA zone 510. Similarly, a second AOI camera 561 may be positioned to capture images of PCBAs within the second PCBA zone 511. and a third AOI camera 562 may be positioned to capture images of PCBAs within the third PCBA zone 512.

[0084] The computer systems associated with the AOI system 500 (as described in FIG. 2) may use the AOI camera imaging and analysis thereof to make a quality offset determination 570. In some cases, the quality offset determination 570 be generated through the analysis of images captured by one or more of the first AOI camera 560, the second AOI camera 561, and the third AOI camera 562 to detect placement defects and calculate accuracy offsets for components placed on PCBAs in each respective PCBA zone.

[0085] Based on the quality offset determination 570, the AOI system 500 may generate offset data feedback for each individual pick and place machine. A first offset data feedback 580 may be provided to the first pick and place machine 501, a second offset data feedback 581 may be provided to the second pick and place machine 502, and a third offset data feedback 582 may be provided to the third pick and place machine 503.

[0086] In some cases, the computer system(s) in the AOI system 500 may be configured to compile and analyze placement accuracy data from multiple PCBAs across the first pick and place machine 501, the second pick and place machine 502, and the third pick and place machine 503 to identify statistical trends in placement errors. This analysis may involve processing data from multiple PCBAs as they move through the board flow through line 550 and are processed by each pick and place machine.

[0087] The computer system(s) in the AOI system 500 may also be configured to perform root cause analysis of placement errors. In some cases, this analysis may be based on the compiled placement accuracy data and machine characteristic data from the first pick and place machine 501. the second pick and place machine 502. and the third pick and place machine 503. The machine characteristic data may include information such as component pick-up success rates, nozzle performance metrics, or other operational parameters specific to each pick and place machine.

[0088] By integrating AOI capabilities across multiple pick and place machines and centralizing data analysis, the AOI system 500 may enable comprehensive monitoring and optimization of the entire SMT production line. The system may allow for dynamic adjustment of placement parameters based on real-time analysis, potentially leading to improved placement accuracy and reduced defect rates across the production process.

[0089] FIG. 6 illustrates a flowchart of a process 600 for component placement and inspection in a pick and place machine, in accordance with one embodiment.

[0090] The process 600 may begin with a step 602. where a baseline image of a printed circuit board assembly (PCBA) may be captured. In some cases, the AOI Camera 106 may capture the baseline image to determine locations of pad sites relative to reference marks on the PCBA 114.

[0091] Following the baseline image capture, the process 600 may proceed to a step 604, where components may be placed on the PCBA 114. The Pick and Place Head 104 may perform this component placement operation.

[0092] After component placement, the process 600 may move to a step 606, where a postplacement image of the PCBA 114 may be captured. The AOI Camera 106 may capture this post-placement image.

[0093] The process 600 may then advance to a step 608, where the post-placement image may be analyzed. In some cases, the Processor 110 may analyze the post-placement image to detect placement defects and calculate accuracy offsets for the placed components.

[0094] Following the analysis, the process 600 may reach a decision point 610, where a determination may be made whether placement defects are detected. If placement defects are detected, the process 600 may proceed along a "Yes" branch to a step 612.

[0095] At the step 612, a message may be presented to a user for defect verification. In some cases, the Processor 110 may generate this message to allow the user to verify detected placement defects as true failures or false failures.

[0096] The process 600 may then move to a step 616, where machine learning may be performed to improve defect detection. In some cases, the Processor 110 may use the user's verification of detected placement defects to enhance defect detection capabilities over time.

[0097] If no placement defects are detected at the decision point 610, the process 600 may follow a "No" branch to a step 614. At the step 614, accuracy offsets may be calculated for the placed components. The Processor 110 may perform these calculations based on the analysis of the post-placement image.

[0098] Following either the step 616 or the step 614, the process 600 may proceed to a step 618. At the step 618, the placement recipe may be modified based on the calculated offsets. In some cases, the Processor 110 may adjust component placement coordinates or other parameters in the placement recipe to improve accuracy in subsequent operations.

[0099] The process 600 may conclude at a step 620, where placement accuracy data may be compiled and analyzed. In some cases, the Processor 110 may compile and analyze placement accuracy data from multiple PCBAs to identify statistical trends in placement errors.

[0100] In some cases, the Processor 1 10 may perform root cause analysis of placement errors. This analysis may be based on the compiled placement accuracy data and machine characteristic data from the SMT Pick and Place Machine 100.

[0101] By integrating image capture, component placement, defect detection, machine learning, and placement recipe modification, the process 600 may enable continuous monitoring and optimization of the SMT manufacturing process. This approach may lead to improved placement accuracy, reduced defect rates, and enhanced overall production efficiency.

[0102] While various examples have been shown and described, the description is intended to be exemplary, rather than limiting and it should be understood by those of ordinary skill in the art that various changes in form and detail may be made therein without departing from the scope of the invention as recited in the accompanying claims.

Claims

CLAIMS1. A pick and place machine, comprising: a gantry system; a pick and place head movable by the gantry system: an automated optical inspection (AOI) camera mounted within the pick and place machine and configured to capture images of an entire printed circuit board assembly (PCBA): and a processor configured to: analyze the images to detect placement defects and calculate accuracy7offsets for components placed on the PCBA, and modify a placement recipe of the pick and place machine based on the calculated accuracy offsets to improve placement accuracy.

2. The pick and place machine of claim 1, wherein the AOI camera is configured to capture a baseline image of the PCBA before component placement and a second image after component placement.

3. The pick and place machine of claim 2, wherein the processor is further configured to: compare the baseline image and the second image to determine locations of components placed relative to reference points on the PCBA.

4. The pick and place machine of claim 1, wherein the processor is further configured to: present a message to a user to verify detected placement defects as true failures or false failures.

5. The pick and place machine of claim 4, wherein the processor is further configured to: perform machine learning to improve defect detection capabilities based on user verification of detected placement defects.

6. The pick and place machine of claim 1, wherein the processor is further configured to: compile and analyze placement accuracy data from multiple PCBAs to identify statistical trends in placement errors.

7. The pick and place machine of claim 6, wherein the processor is further configured to:perform root cause analysis of placement errors based on the compiled placement accuracy data and machine characteristic data.

8. A method of operating a pick and place machine, comprising: capturing, using an automated optical inspection (AOI) camera mounted within the pick and place machine, a baseline image of a printed circuit board assembly (PCBA) to determine locations of pad sites relative to reference marks; placing components on the PCBA; capturing, using the AOI camera, a post-placement image of the PCBA; analyzing the post-placement image to detect placement defects and calculate accuracy offsets for the placed components; and modify ing a placement recipe of the pick and place machine based on the calculated accuracy offsets.

9. The method of claim 8, further comprising: presenting a message to a user to verify detected placement defects as true failures or false failures.

10. The method of claim 9, further comprising: performing machine learning to improve defect detection capabilities based on user verification of detected placement defects.

11. The method of claim 8, further comprising: compiling and analyzing placement accuracy data from multiple PCBAs to identify statistical trends in placement errors.

12. The method of claim 11, further comprising: performing root cause analysis of placement errors based on the compiled placement accuracy data and machine characteristic data.

13. The method of claim 8, wherein analyzing the post-placement image comprises: comparing the baseline image and the post-placement image to determine locations of components placed relative to the reference marks on the PCBA.

14. The method of claim 13, wherein modifying the placement recipe comprises: adjusting component placement coordinates based on the determined locations of components relative to the reference marks.

15. A system for improving placement accuracy in printed circuit board assembly manufacturing, comprising: a plurality of pick and place machines arranged in a production line; an automated optical inspection (AOI) camera mounted within at least one of the pick and place machines; and a processor configured to: analyze images captured by the AOI camera to detect placement defects and calculate accuracy offsets for components placed on printed circuit board assemblies (PCBAs), and distribute placement recipe modifications based on the calculated accuracy offsets to the plurality of pick and place machines.

16. The system of claim 15, wherein the AOI camera is configured to capture a baseline image of each PCBA before component placement and a post-placement image after component placement.

17. The system of claim 16, wherein the processor is further configured to: compare the baseline image and the post-placement image to determine locations of components placed relative to reference marks on each PCBA.

18. The system of claim 17, wherein the processor is further configured to: adjust component placement coordinates in the placement recipe modifications based on the determined locations of components relative to the reference marks.

19. The system of claim 15, wherein the processor is further configured to: compile and analyze placement accuracy data from multiple PCBAs across the plurality of pick and place machines to identify statistical trends in placement errors.

20. The system of claim 19, wherein the processor is further configured to: perform root cause analysis of placement errors based on the compiled placement accuracy data and machine characteristic data from each of the plurality of pick and place machines.

Images