Handheld laser tool sensor system

Abstract

A handheld laser tool system comprises a handheld laser tool (106) with a body portion (114), laser beam pathway, and forward laser beam exit (115). The system includes a sensor suite (196) with an image sensor (173) detecting laser light reflected from a workpiece (312) and a processing circuit (182) determining distance between the tool (106) and workpiece (312) using triangulation. A controller (180) adjusts operational parameters of the handheld laser tool (106) based on the determined distance. The sensor suite (196) may include an inertial measurement unit (304) detecting orientation and movement of the tool (106), where the processing circuit (182) performs sensor fusion between image sensor (173) and inertial measurement unit (304) data to determine three-dimensional position of the handheld laser tool (106) relative to the workpiece (312).

Description

Attorney Docket No. 5117.0025W01HANDHELD LASER TOOL SENSOR SYSTEMCROSS REFERENCE TO RELATED APPLICATIONS

[0001] This patent application claims the benefit of U.S. Provisional Application No. 63 / 707035, titled MODULAR HANDHELD LASER MATERIAL PROCESSING SYSTEM, filed October 14, 2024, the contents of which are expressly incorporated by reference herein.TECHNICAL FIELD

[0002] The present disclosure relates to handheld laser material processing systems that provide real-time monitoring and control of operational parameters through distance measurement, proximity detection, and sensor fusion techniques.BACKGROUND

[0003] Handheld laser material processing systems have gained popularity due to their portability and precision in various applications including welding, cutting, and cleaning operations. These systems typically consist of a portable laser radiation generation and control unit connected to a handheld laser tool through an umbilical cord containing optical fibers, electrical connections, and process gas lines. As laser technology has advanced and costs have decreased, these handheld systems have become more accessible to smaller machine shops and individual users.

[0004] However, existing handheld laser systems face challenges in maintaining optimal operational parameters during material processing operations. Traditional systems often rely on electrical contact between the tool tip and workpiece for proximity detection and safety control, which can be problematic when working with nonconductive materials, coated surfaces, or when traversing gaps in the workpiece. Additionally, maintaining proper focus distance and tool orientation relative to the workpiece surface can be difficult during handheld operations, potentially affecting processing quality and consistency. The lack of real-time feedback regarding tool position, workpiece proximity, and processing conditions can limit the effectiveness and safety of handheld laser material processing operations.SUMMARY

[0005] In some examples, the disclosure describes a handheld laser tool system that includes a handheld laser tool having a body portion with a laser beam pathway and a forward laser beam exit. The system includes a sensor suite operatively coupled to theAttorney Docket No. 5117.0025W01 handheld laser tool, where the sensor suite includes an image sensor configured to detect laser light reflected from a workpiece and a processing circuit configured to determine a distance between the handheld laser tool and the workpiece based on triangulation using the detected laser light. The system includes a controller operatively coupled to the sensor suite and configured to adjust at least one operational parameter of the handheld laser tool based on the determined distance.

[0006] In some examples, the disclosure describes a method of controlling a handheld laser tool during material processing that includes directing laser light from the handheld laser tool toward a workpiece. The method includes detecting, with an image sensor, laser light reflected from the workpiece. The method includes determining, with a processing circuit, a distance between the handheld laser tool and the workpiece based on triangulation using the detected laser light. The method includes automatically adjusting at least one operational parameter of the handheld laser tool based on the determined distance.

[0007] In some examples, the disclosure describes a handheld laser tool system that includes a handheld laser tool having a body portion with a laser beam pathway and a forward laser beam exit. The system includes a capacitance sensor electrically coupled to a distal tip of the handheld laser tool and configured to detect a capacitance between the distal tip and a workpiece. The system includes a controller operatively coupled to the capacitance sensor and configured to determine a proximity of the distal tip to the workpiece based on the detected capacitance and to control laser emission based on the determined proximity.

[0008] A feature and advantage of the present disclosure is that the triangulation-based distance measurement system enables accurate positioning of the handheld laser tool relative to the workpiece without requiring physical contact, thereby allowing uninterrupted material processing operations over gaps, corners, and nonconductive materials.

[0009] Feature and advantages of the present disclosure including, but are not limited to: (1) the sensor fusion between image sensor data and inertial measurement unit data provides three-dimensional position determination of the handheld laser tool relative to the workpiece, enabling precise control of operational parameters during handheld operations; (2) the automatic adjustment of multiple operational parameters including laser power, process gas flow rate, and wobble frequency based on determined distance improves material processing quality and consistency; (3) the acoustic sensor correlation with distance measurements enables real-time assessment of material processing quality through acoustic signature analysis; and (4) the system achieves distance measurement resolution of 200 micrometers or better, providing precise control for high-quality material processing operations.Attorney Docket No. 5117.0025W01BRIEF DESCRIPTION OF FIGURES

[0010] FIGS. 1A and IB are conceptual diagrams illustrating an example modular handheld laser system and controller architecture with sensor suite.

[0011] FIG. 2A is a conceptual diagram illustrating an example laser triangulation schematic showing geometric relationships between system components.

[0012] FIG. 2B is a conceptual diagram illustrating an example laser triangulation measurement system with camera sensor and surface detection.

[0013] FIG. 3 is a conceptual diagram illustrating an example experimental apparatus with torch, sensors, and workpiece mounting arrangement.

[0014] FIGS. 4A and 4B are conceptual diagrams illustrating example photographs of torch tip positioning relative to workpiece surface.

[0015] FIG. 5 is a conceptual diagram illustrating an example graph of image sensor readings versus gap distance measurements.

[0016] FIG. 6 is a conceptual diagram illustrating an example graph of image sensor readings versus marker laser position.

[0017] FIG. 7 is a conceptual diagram illustrating an example side view of torch with integrated sensor module and optical axes.

[0018] FIG. 8 is a conceptual diagram illustrating an example side view of torch with dual camera sensor configuration.

[0019] FIG. 9 is a conceptual diagram illustrating an example set of graphs showing position estimation using Kalman filter processing.

[0020] FIG. 10 is a conceptual diagram illustrating an example graph of angular step estimation measurements over time.

[0021] FIGS. 11 and 12 are conceptual diagrams illustrating example graphs of spectrometer output versus wavelength at 200W laser power and 500W laser power, respectively.

[0022] FIG. 13 is a conceptual diagram illustrating an example graph of spectrometer output versus wavelength at 800W laser power.

[0023] FIG. 14 is a conceptual diagram illustrating an example experimental apparatus with torch and fiber optic spectrometer configuration.

[0024] FIG. 15 is a conceptual diagram illustrating an example confusion matrix for power and gap distance combinations.

[0025] FIG. 16 is a conceptual diagram illustrating an example photograph of weld penetration depth cross-section measurements.Attorney Docket No. 5117.0025W01

[0026] FIG. 17 is a conceptual diagram illustrating an example graph of predicted versus true central depth measurements.

[0027] FIG. 18 is a conceptual diagram illustrating an example graph of predicted versus true maximum depth measurements.

[0028] FIG. 19 is a conceptual diagram illustrating an example graph of feature importance analysis for frequency ranges.DETAILED DESCRIPTION

[0029] The following description sets forth exemplary aspects of the present disclosure. It should be recognized, however, that such description is not intended as a limitation on the scope of the present disclosure. Rather, the description also encompasses combinations and modifications to those exemplary aspects described herein.

[0030] FIG. 1A illustrates an example modular handheld laser system 100 configured for material processing operations. Modular handheld laser system 100 may include a base unit 104, a torch 106, and an umbilical cord 108 connecting base unit 104 and torch 106. Torch 106 may be connected to base unit 104 at a universal receptacle 110 positioned at a handgrip 112 of torch 106. Handgrip 112 is joined to a body 114 which defines a body portion having a laser beam pathway extending therethrough and terminating at a forward laser beam exit at a laser discharge end 115. A nozzle 116 may be removably attached at the laser discharge end 115 to direct laser energy toward a target surface during material processing operations.

[0031] Base unit 104 may house various components for generating and controlling laser energy. A laser generation source 120 may be positioned within base unit 104 to generate laser radiation for transmission through umbilical cord 108 to torch 106. A controller 121 may be operatively coupled to the laser generation source 120 to control laser output parameters. Base unit 104 may further include a coolant system 122 for thermal management and a process gas control system 123 for supplying process gases to torch 106 during operation.

[0032] A user interface 125 may be provided on base unit 104 in the form of controls, indicators, touchscreen displays, or combinations thereof. User interface 125 may allow operators to adjust operational parameters and monitor system status during material processing operations.

[0033] Modular handheld laser system 100 may include electrical connections for safety and operational control. A conductive wire 126 may connect a workpiece 124 to base unit 104, creating a complete electrical circuit when nozzle 116 contacts workpiece 124. ThisAttorney Docket No. 5117.0025W01 electrical connection may provide assurance that laser energy is directed toward the intended workpiece 124.

[0034] Torch 106 may include user-operable controls for material processing operations. A first trigger 113A and a second trigger 113B may be positioned on handgrip 112 to allow operators to control laser activation and operational parameters. First trigger 113A and second trigger 113B may be configured as single-stage or multi-stage triggers, providing operators with precise control over laser output during material processing operations.

[0035] Referring to FIG. IB, a controller 180 may be operatively coupled to modular handheld laser system 100 to provide control and monitoring capabilities. Controller 180 may include processing circuitry 182 configured to execute control algorithms and process sensor data. A memory 184 may be operatively coupled to processing circuitry 182 to store operational parameters, sensor data, and control instructions. An operations module 186 may be configured to control various operational parameters of modular handheld laser system 100 based on inputs from sensors and user commands. Communications circuitry 188 may be provided to enable data transmission between controller 180 and other system components.

[0036] Controller 180 may be operatively coupled to a control unit 192 and a remote device 194 through the communications circuitry 188. The control unit 192 may provide centralized control functions for modular handheld laser system 100, while the remote device 194 may allow remote monitoring and control of system operations. In some examples, control unit 192 may be collocated on base unit 104.

[0037] A sensor suite 196 may be operatively coupled to controller 180 to monitor various aspects of material processing operations. The sensor suite 196 may include an image sensor 173 configured to detect laser light reflected from workpiece 124 or plasma produced at workpiece 124. Processing circuitry 182 may be configured to determine a distance between torch 106 and workpiece 124 based on triangulation using the detected light from image sensor 173. Controller 180 may be configured to adjust at least one operational parameter of torch 106 based on the determined distance.

[0038] Sensor suite 196 may further include a photodiode 161 configured to detect laser radiation scattered by contamination on protective windows or other optical components. A thermosensor 163 may be positioned to monitor temperature conditions within torch 106 or at material processing locations. A plasma sensor 165 may be configured to detect plasma generated during material processing operations, providing feedback on process conditions and quality. In some examples, plasma sensor 165 may include the plasma sensors, and torch 106 may include the handheld laser system including controllers, as described in U.S. PatentAttorney Docket No. 5117.0025W01Application Publication No. 2024 / 0009758, the entirety of which is incorporated by reference herein.

[0039] An acoustic sensor 167 may be included in the sensor suite 196 to detect sound generated during material processing operations. Acoustic sensor 167 may be configured to monitor acoustic signatures that correlate with material processing quality and operational parameters. A back reflection sensor 169 may be positioned to detect laser radiation reflected back from the workpiece during material processing, providing information about process penetration and quality.

[0040] A capacitance sensor 171 may be electrically coupled to a distal tip of torch 106 and configured to detect a capacitance between the distal tip and workpiece 124. Controller 180 may be operatively coupled to capacitance sensor 171 and configured to determine a proximity of the distal tip to workpiece 124 based on the detected capacitance and to control laser emission based on the determined proximity.

[0041] Sensor suite 196 may include a spectrometer 175 for laser-induced breakdown spectroscopy (LIBS) to identify workpiece material composition. Spectrometer 175 may analyze light emitted by plasma created during laser processing to determine elemental composition of the workpiece material. In some examples, the LIBS system and method may include any suitable LIBS system, such as the LIBS system and method described in U.S. Patent No. 10,871,450, the entirety of which in incorporated by reference herein. In some cases, sensor suite 196 may include a multi-channel spectral sensor with 14 channels covering 380 nm to 1000 nm wavelength range for spectral analysis of material processing operations. Sensor suite 196 may include a volatile organic compound (VOC) sensor for detecting hazardous gas emissions during material processing operations.

[0042] An input suite 198 may be operatively coupled to controller 180 to receive user inputs and system status information. Input suite 198 may include a nozzle detector 103 configured to identify the type of nozzle 116 attached to torch 106. A microphone 105 may be provided to receive voice commands from operators for controlling system parameters. A torch detector 111 may be configured to identify the type of torch 106 connected to the system, allowing controller 180 to adjust operational parameters accordingly.

[0043] Input suite 198 may further include a mode switch 118 positioned on torch 106 to allow operators to select different operational modes during material processing. A display 119, which may include a touch screen display, may be provided on torch 106 to show operational parameters, system status, and other relevant information to operators.Attorney Docket No. 5117.0025W01

[0044] An output suite 199 may be operatively coupled to controller 180 to provide feedback to operators. Output suite 199 may include display 119; a speaker 117 configured to provide audible alerts, confirmations, and voice prompts to operators; and / or a vibration motor 127 positioned within torch 106 to provide haptic feedback in response to operational changes, voice commands, or system alerts.

[0045] FIG. 2A is a laser triangulation schematic 200 illustrating the geometric relationships between various components of a distance measurement system for modular handheld laser system 100. The laser triangulation schematic 200 may demonstrate how triangulation principles may be applied to determine distances between torch 106 and a workpiece during material processing operations.

[0046] For example, a coordinate of IMU sensor 202 provides a reference frame for orientation and movement measurements relative to a global coordinate system of a world frame 210. An image sensor 204 positioned along a plane 205 is configured to detect laser light reflected from the workpiece. Image sensor lens 206 that defines an image sensor optical axis 208 extending from the image sensor 204 toward the workpiece. Laser vector 211 represents the direction and path of laser light projected toward the workpiece. The geometric arrangement between laser vector 211 and image sensor optical axis 208 may create a mutual angle 212 between laser vector 211 and image sensor optical axis 208. The triangulationbased distance measurement is based on mutual angle 212, which is constant.

[0047] In some cases, mutual angle 212 between laser vector 211 and image sensor optical axis 208 may be between 10 degrees and 45 degrees to provide suitable triangulation geometry for accurate distance measurements. In other examples, mutual angle 212 may be between about 1 degrees and 10 degrees, such as less than 5 degrees or less than 4 degrees. Mutual angle 212 may be selected to optimize measurement sensitivity while minimizing shadowing effects and maintaining adequate measurement range.

[0048] A torch orientation 214 illustrates the angular position of torch 106 during material processing operations, which varies based on a position of torch 106 relative to workpiece 124. The torch orientation 214 may affect the geometric relationships within the triangulation system and may be monitored by the IMU to maintain accurate distance measurements.

[0049] The laser triangulation schematic 200 may include vertical measurement parameters for distance determination. A vertical focal height 216 may represent the vertical distance component within the triangulation geometry. A focus distance 218 may be measured between focal planes to establish the working distance of the laser system. AnAttorney Docket No. 5117.0025W01 upper focal plane 220 and a lower focal plane 222 may define the boundaries of the measurement range, with the focus distance 218 extending between the upper focal plane 220 and the lower focal plane 222.

[0050] The geometric relationships between torch orientation 214 and mutual angle 212 may enable precise triangulation-based distance measurements. Processing circuitry 182 may utilize these geometric relationships to calculate the distance between torch 106 and the workpiece based on the position of reflected laser light detected by the image sensor 204. The vertical focal height 216, focus distance 218, upper focal plane 220, and lower focal plane 222 may provide reference parameters for determining the working distance and maintaining proper focus during material processing operations.

[0051] Referring to FIG. 2B, a laser triangulation schematic 250 may illustrate the fundamental measurement principles used by modular handheld laser system 100 for determining distances between torch 106 and workpiece 124. For example, a laser 252 may project laser light toward a surface 258 (i.e., workpiece 124), creating a detectable light pattern. A camera sensor 254 may be positioned at a mutual angle 264 relative to a laser vector of laser 252 to detect reflected laser light from the surface 258. A camera lens 256 with a known magnification (m) may be positioned in front of the camera sensor 254 to focus reflected laser light onto the camera sensor 254.

[0052] Mutual angle 264 may define the triangulation angle that enables distance calculations based on the position of reflected laser light detected by the camera sensor 254. The triangulation measurement principle may be based on the mathematical relationship between surface movement and corresponding changes in the detected light pattern position. For example, a light-pattern translation on camera sensor (Av) 260 may occur when the surface 258 moves vertically relative to the laser 252 and camera sensor 254. The mathematical relationship between the light-pattern translation on camera sensor (Av) 260 and the vertical movement of the surface (Az) 262 may be expressed as Av = Az • m • sin a, where m represents the optical magnification of the camera lens with magnification m 256 and a represents the mutual angle 264 between the light plane and the camera’s optical axis (a). This mathematical relationship may enable processing circuitry 182 to calculate precise distance measurements based on detected changes in light pattern position on the camera sensor 254.

[0053] Processing circuitry 182 may be configured to determine a distance between torch 106 and the workpiece based on triangulation using the detected laser light from image sensor 173. The triangulation calculations may utilize the mathematical relationship Av = AzAttorney Docket No. 5117.0025W01• m • sin a to convert detected light-pattern translation on camera sensor (Av) 260 into corresponding vertical movement of the surface (Az) 262 measurements.

[0054] The selection of optical magnification m and the mutual angle 264 between the light plane and the camera’s optical axis (a) may affect the measurement sensitivity and range of the triangulation system. Higher optical magnification m values may increase measurement sensitivity but may reduce the measurement range due to field of view limitations. Larger values for the mutual angle 264 between the light plane and the camera’s optical axis (a) may increase measurement sensitivity but may introduce shadowing effects that limit measurement accuracy in certain geometric configurations and / or present drawbacks relative to handheld system design and operation.

[0055] In some cases, the optical magnification m may be selected to be less than 0.1 to provide adequate measurement range while maintaining suitable sensitivity for material processing applications. The mutual angle 264 between the light plane and the camera’s optical axis (a) may be selected to be less than 30 degrees to minimize shadowing effects while providing adequate triangulation sensitivity.

[0056] Processing circuitry 182 may be configured to determine the distance with a resolution of 200 micrometers or better using the triangulation principles illustrated in the laser triangulation schematics 200 and 250. The resolution capability may be achieved through the mathematical relationship between the light-pattern translation on camera sensor (Av) 260, the vertical movement of the surface (Az) 262, the optical magnification m, and the mutual angle 264 between the light plane and the camera’s optical axis (a). The high resolution distance measurements may enable precise control of material processing operations and accurate positioning of torch 106 relative to the surface 258.

[0057] Referring to FIG. 3, an experimental apparatus 300 demonstrated and validated the distance measurement and sensor fusion capabilities of modular handheld laser system 100. Experimental apparatus 300 included a controlled testing environment for evaluating triangulation-based distance measurements, orientation tracking, and acoustic monitoring during material processing operations.

[0058] A torch 302 was mounted on a torch mount 308 for stable positioning and precise control over the torch 302 location relative to a workpiece 312 on a workpiece mount 314. An IMU 304 was integrated with torch 302 to provide orientation and movement data during experimental operations. A camera 306 was positioned near torch 302 to capture images of laser light reflected from workpiece 312 during experimental operations to enable triangulation-based distance calculations. A microphone 310 configured to detect soundAttorney Docket No. 5117.0025W01 generated during laser welding, cutting, or cleaning operations, e.g., in frequency ranges from 100 Hz to 80 kHz with frequencies greater than 20kHz being beyond human hearing, monitored acoustic signatures generated during material processing operations.

[0059] Workpiece 312 was represent various materials that may be processed using modular handheld laser system 100, including metals, alloys, and other materials suitable for laser welding, cutting, or cleaning operations. The workpiece mount 314 allowed for precise positioning and controlled movement of workpiece 312 to test various distance measurement scenarios and geometric configurations. The arrangement of torch 302, IMU 304, camera 306, and microphone 310 enabled simultaneous collection of multiple sensor data streams during material processing operations. Processing circuitry 182 received and processed data from the IMU 304, camera 306, and microphone 310 to perform sensor fusion calculations that enhanced distance measurement accuracy and three-dimensional position determination of torch 302 relative to workpiece 312, providing comprehensive monitoring of material processing operations and validating the effectiveness of sensor fusion algorithms for three- dimensional position determination and material processing quality assessment, including penetration depth, weld quality, and process stability.

[0060] FIGS. 4A and 4B are photographs illustrating how torch 106 may be positioned relative to workpiece surfaces and how marker laser beams may be projected onto workpieces for triangulation-based distance measurement. A photo 402 of a torch tip 410 and a workpiece 412 shown in FIG. 4A, depicts torch tip 410 positioned near workpiece 412 during material processing operations. A photo 404 of torch tip 410 with a marker laser spot 414 on workpiece 412 is shown in FIG. 4B, illustrating how the marker laser beam creates a marker laser spot 414 on workpiece 412, e.g., that may be detected by image sensor 173 for triangulation-based distance calculations. The relative size, pattern, and / or intensity of marker laser spot 414 may be used, optionally with additional data such as IMU data, to determine the triangulation-based distance calculations.

[0061] FIG. 5 is a graph of image sensor reading versus gap distance 500 that demonstrates the linear relationship between sensor measurements and distance variations during material processing operations as determined using experimental apparatus 300 with a 200W infrared laser on a stainless steel coupon. Image sensor readings were plotted in pixels on the vertical axis against gap distance measurements plotted in micrometers on the horizontal axis. The data points in the graph of image sensor reading versus gap distance 500 demonstrated a linear relationship with a negative slope of approximately -0.1041, indicating that the image sensor reading decreases as the gap distance increases between torch 106 andAttorney Docket No. 5117.0025W01 workpiece 312. The linear relationship shown in the graph of image sensor reading versus gap distance 500 may enable processing circuitry 182 to perform accurate distance calculations based on detected changes in pixel-based image sensor readings.

[0062] The data presented in the graph of image sensor reading versus gap distance 500 may demonstrate the capability of modular handheld laser system 100 to track distance variations with high resolution during material processing operations. The graph of image sensor reading versus gap distance 500 shows distance tracking with 200 micrometer steps, demonstrating the sensor’s ability to detect workpiece proximity changes with precision suitable for material processing control applications. The high resolution capability demonstrated in the graph of image sensor reading versus gap distance 500 may enable processing circuitry 182 to achieve distance measurement resolution of 50 micrometers or better using marker laser triangulation to support fine- scale distance control during material processing operations.

[0063] A method of controlling torch 106 during material processing may include detecting, with image sensor 173, laser light reflected from workpiece 312, as demonstrated by the sensor response data in the graph of image sensor reading versus gap distance 500. The method may further include determining, with processing circuitry 182, a distance between torch 106 and workpiece 312 based on triangulation using the detected laser light, utilizing the linear relationship shown in the graph of image sensor reading versus gap distance 500 to convert sensor readings into accurate distance measurements.

[0064] FIG. 6 is a graph of image sensor reading in pixels versus marker laser position 600 on a stainless steel coupon at a gap distance of 50 micrometers. Graph 600 demonstrates how image sensor 173 responds to controlled changes in marker laser spot position on workpiece 312, providing validation of the triangulation measurement system’s tracking accuracy. The data points provide a positive linear relationship with the sensor readings increasing from approximately 500 pixels at position 0 to approximately 700 pixels at position 3000 micrometers. The linear relationship enables processing circuitry 182 to perform precise marker laser spot position calculations based on detected pixel changes in image sensor 173. Additionally, the experimental data presented in graph 600 demonstrate marker laser spot tracking using 50 micrometer position steps, showing the capability of modular handheld laser system 100 to detect small position changes with high precision.

[0065] A method of controlling torch 106 during material processing may include directing laser light from the handheld laser tool toward workpiece 312, where the laser light may include a marker laser beam that creates a detectable spot on workpiece 312. The markerAttorney Docket No. 5117.0025W01 laser beam projection may enable image sensor 173 to track spot position changes as demonstrated in the graph of image sensor reading versus marker laser position 600, providing the optical reference needed for triangulation-based distance calculations during material processing operations.

[0066] The positive linear relationship shown in the graph of image sensor reading versus marker laser position 600 may complement the negative linear relationship demonstrated in the graph of image sensor reading versus gap distance 500, providing comprehensive validation of the triangulation measurement system’s accuracy across different measurement scenarios. The combination of position tracking capability shown in the graph of image sensor reading versus marker laser position 600 and distance measurement capability may enable processing circuitry 182 to perform precise three-dimensional positioning calculations for torch 106 relative to workpiece 312 during material processing operations.

[0067] FIG. 7 illustrates an example torch 700 configured with an integrated sensor module 708 positioned to provide comprehensive monitoring and distance measurement capabilities during material processing operations. The torch 700 may include a torch tip 702 positioned at a distal end of the torch 700 for directing laser energy toward target workpieces. A laser axis 704 may be defined within the torch 700 to establish the primary optical path for laser energy transmission from the torch 700 to the workpiece surface.

[0068] The sensor module 708 may be positioned in a line of sight of a beam focal point of the torch 700 to enable direct monitoring of material processing operations. The sensor module 708 may be configured to house multiple sensing components that work together to provide distance measurement, orientation tracking, and acoustic monitoring capabilities during material processing operations.

[0069] A camera optical axis 706 may be defined by optical components within the sensor module 708, where the camera optical axis 706 intersects the laser axis 704 to create the geometric configuration needed for triangulation-based distance measurements. The intersection between the camera optical axis 706 and the laser axis 704 may establish the mutual angle between the laser vector and the image sensor optical axis that enables precise triangulation calculations for determining distances between the torch 700 and workpiece surfaces.

[0070] A camera 710 may be positioned within the sensor module 708 along the camera optical axis 706 to detect laser light reflected from workpiece surfaces during material processing operations. The camera 710 may be configured to capture images of marker laser spots projected onto workpiece surfaces, enabling processing circuitry 182 to performAttorney Docket No. 5117.0025W01 triangulation calculations based on detected light pattern positions. The camera 710 may work in conjunction with other sensors within the sensor module 708 to provide comprehensive spatial positioning data for the torch 700 relative to target workpieces.

[0071] An IMU sensor 712 may be integrated within the sensor module 708 to detect orientation and movement of the torch 700 during material processing operations. The IMU sensor 712 may be configured to provide three-axis acceleration and three-axis gyroscopic measurements that enable tracking of torch 700 position and angular orientation changes. Processing circuitry 182 may be configured to perform sensor fusion between data from image sensor 173 and data from the IMU sensor 712 to determine a three-dimensional position of the torch 700 relative to the workpiece.

[0072] The sensor fusion capabilities may combine triangulation-based distance measurements from the camera 710 with orientation and movement data from the IMU sensor 712 to provide enhanced spatial positioning accuracy. Processing circuitry 182 may utilize sensor fusion algorithms to integrate the different sensor data streams and calculate comprehensive three-dimensional position information that accounts for both distance and angular positioning of the torch 700 relative to workpiece surfaces.

[0073] A bandpass filter 714 may be positioned within the sensor module 708 to enhance the detection capabilities of the camera 710 for marker laser identification. The bandpass filter 714 may be centered at 650 nm wavelength to provide selective transmission of marker laser light while filtering out other wavelengths that may interfere with distance measurement operations. The bandpass filter 714 may enable the camera 710 to distinguish marker laser light from ambient lighting conditions and other optical sources that may be present during material processing operations.

[0074] The bandpass filter 714 centered at 650 nm wavelength may correspond to the wavelength of marker laser beams used for triangulation-based distance measurements. The selective wavelength transmission provided by the bandpass filter 714 may improve the signal-to-noise ratio of marker laser spot detection, enabling more accurate triangulation calculations and enhanced distance measurement precision during material processing operations.

[0075] An ultrasonic microphone module 716 may be integrated within the sensor module 708 to monitor acoustic signatures generated during material processing operations. The ultrasonic microphone module 716 may be configured to operate in frequency ranges from 100 Hz to 80 kHz, providing enhanced material processing monitoring capabilities beyond conventional audio frequency ranges. The extended frequency range of the ultrasonicAttorney Docket No. 5117.0025W01 microphone module 716 may enable detection of acoustic signatures that correlate with material processing quality and operational parameters.

[0076] Processing circuitry 182 may be configured to correlate acoustic signatures from the ultrasonic microphone module 716 with the determined distance to assess material processing quality. The correlation between acoustic signatures and distance measurements may provide comprehensive feedback on material processing conditions, enabling controller 180 to automatically adjust operational parameters based on both spatial positioning and acoustic quality indicators.

[0077] The camera 710 exposure timing may be synchronized to laser pulse timing to capture workpiece surface images with and without laser marker illumination. The synchronized exposure timing may enable the camera 710 to capture reference images of the workpiece surface without marker laser illumination and comparison images with marker laser spots, providing enhanced contrast and detection accuracy for triangulation calculations. The synchronized exposure timing may also enable the camera 710 to capture images during specific phases of pulsed laser operations, providing temporal correlation between distance measurements and material processing events.

[0078] The combination of the camera 710, IMU sensor 712, bandpass filter 714, and ultrasonic microphone module 716 within the sensor module 708 may provide comprehensive monitoring capabilities for material processing operations. The sensor module 708 may enable simultaneous collection of optical, inertial, and acoustic data streams that may be processed by processing circuitry 182 to provide real-time feedback and control for material processing operations.

[0079] Controller 180 may be configured to automatically adjust at least one operational parameter of the torch 700 based on the determined distance from triangulation calculations and the comprehensive sensor data from the sensor module 708. The operational parameter adjustments may include laser power output, process gas flow rates, wobble frequency, and other processing parameters that affect material processing quality and efficiency. The automatic parameter adjustment capabilities may enable responsive control of material processing operations based on real-time distance measurements and sensor feedback from the sensor module 708.

[0080] FIG. 8 illustrates an example torch 800 configured with a dual camera sensor module 808 that provides enhanced optical detection capabilities for material processing operations. The torch 800 may include a torch tip 802 positioned at a distal end for directing laser energy toward target workpieces during material processing operations. A laser axis 804Attorney Docket No. 5117.0025W01 may be defined within the torch 800 to establish the primary optical path for laser energy transmission from the torch 800 to workpiece surfaces.

[0081] A camera optical axis 806 may be defined by optical components within the sensor module 808, where the camera optical axis 806 intersects the laser axis 804 to create the geometric configuration needed for triangulation-based distance measurements. The intersection between the camera optical axis 806 and the laser axis 804 may establish the mutual angle between the laser vector and the image sensor optical axis that enables precise triangulation calculations for determining distances between the torch 800 and workpiece surfaces.

[0082] The sensor module 808 may be positioned to monitor torch 800 operations through a dual camera deployment configuration that provides enhanced optical detection capabilities. The dual camera deployment within the sensor module 808 may enable simultaneous capture of different optical wavelengths and imaging modes during material processing operations, providing comprehensive optical feedback for distance measurement and process monitoring.

[0083] A mono camera 818 may be positioned within the sensor module 808 along the camera optical axis 806 to detect infrared laser light reflected from workpiece surfaces during material processing operations. The mono camera 818 may be configured to capture reflected infrared laser light at 1064 nm wavelength, enabling distance measurements during actual material processing operations when the primary laser beam is active. The capability of the mono camera 818 to detect infrared laser light at 1064 nm wavelength may allow triangulation measurements using the primary laser beam in addition to dedicated marker laser beams.

[0084] An AR coated window 820 may be positioned between optical components within the sensor module 808 to provide wavelength-selective transmission characteristics. The AR coated window 820 may have highest transmission at 1064 nm wavelength, corresponding to the wavelength of infrared laser beams used for material processing operations. The AR coated window 820 may be configured to transmit most of the 1064 nm infrared laser light while reflecting a small portion of the infrared light toward the mono camera 818 for detection and triangulation calculations.

[0085] An RGB camera 822 may be positioned within the sensor module 808 to capture surface images with marker laser illumination during material processing operations. The RGB camera 822 may be separated from the mono camera 818 by the AR coated window 820, creating a dual camera configuration that enables simultaneous capture of infrared laserAttorney Docket No. 5117.0025W01 light and visible spectrum images. The RGB camera 822 may be configured to capture surface images that include marker laser spots projected onto workpiece surfaces, providing visual documentation and additional optical reference points for material processing monitoring.

[0086] The dual camera deployment configuration may enable the mono camera 818 to capture reflected infrared laser light while the RGB camera 822 captures surface images with marker laser illumination simultaneously during material processing operations. The AR coated window 820 positioned between the mono camera 818 and the RGB camera 822 may provide wavelength separation that allows each camera to operate optimally within its designated spectral range without interference from the other camera’s optical detection requirements.

[0087] The RGB camera 822 exposure timing may be synchronized to marker laser blinking to provide images with and without marker laser beam spot illumination. The synchronized exposure timing may enable the RGB camera 822 to capture reference images of workpiece surfaces without marker laser illumination and comparison images with marker laser spots, providing enhanced contrast and detection accuracy for visual monitoring of material processing operations. The synchronized exposure timing may also enable temporal correlation between surface imaging and marker laser projection events.

[0088] Processing circuitry 182 may be configured to perform sensor fusion between data from image sensor 173 and data from the IMU sensor 712 to determine a three-dimensional position of the torch 800 relative to workpiece surfaces. The sensor fusion may utilize a Kalman filter algorithm to combine the image sensor data and the IMU data, providing enhanced spatial positioning accuracy through mathematical integration of multiple sensor data streams. The Kalman filter algorithm may process triangulation-based distance measurements from the mono camera 818 and orientation data from the IMU sensor 712 to calculate comprehensive three-dimensional position information that accounts for both distance and angular positioning of the torch 800 relative to workpiece surfaces.

[0089] The Kalman filter algorithm may incorporate bias estimation for IMU sensor 712 measurements to minimize position measurement drift during material processing operations. The bias estimation capabilities of the Kalman filter algorithm may enable accurate position tracking even during extended material processing operations where IMU sensor 712 drift might otherwise accumulate and affect measurement accuracy. The Kalman filter algorithm may continuously update position estimates based on new sensor data from both the monoAttorney Docket No. 5117.0025W01 camera 818 and the IMU sensor 712, providing real-time three-dimensional positioning feedback for material processing control.

[0090] The dual camera configuration within the sensor module 808 may provide comprehensive optical monitoring capabilities that complement the sensor fusion algorithms. The mono camera 818 may provide infrared laser light detection for triangulation calculations while the RGB camera 822 may provide visual documentation and marker laser spot tracking, creating multiple optical data streams that may be integrated through the Kalman filter algorithm to enhance overall measurement accuracy and reliability during material processing operations.

[0091] Referring to FIG. 9, graph 900 illustrates position estimation based on acceleration sensor data using Kalman filter and demonstrates the capability of modular handheld laser system 100 to track torch 106 movement through sensor fusion algorithms during material processing operations. Graph 900 illustrates three distinct data streams plotted against time in seconds, including measured acceleration, velocity estimation, and position estimation data that collectively illustrate how the IMU sensor 712 data may be processed to determine spatial positioning of torch 106.

[0092] The measured acceleration data exhibited distinct peaks and valleys between approximately 5-15 seconds, indicating periods of torch 106 acceleration and deceleration during material processing operations. The velocity estimation data provided corresponding positive and negative velocity changes during the same 5-15 second time period where acceleration peaks and valleys occur. The velocity estimation may be calculated by processing circuitry 182 through integration of the measured acceleration data, where the Kalman filter algorithm processes the acceleration measurements to determine velocity changes while accounting for sensor bias and measurement uncertainties. The position estimation data provided a cumulative change in position over the measurement period, with a peak displacement of approximately 5 mm. The position estimation may be calculated by processing circuitry 182 through double integration of the acceleration data, where the Kalman filter algorithm processes the velocity estimates to determine cumulative position changes while minimizing drift effects that may accumulate during the integration process.

[0093] The method of controlling torch 106 during material processing may further include performing sensor fusion between data from image sensor 173 and data from the IMU sensor 712 to determine a three-dimensional position of torch 106 relative to workpiece 312. The sensor fusion capabilities may be demonstrated by the position tracking accuracyAttorney Docket No. 5117.0025W01 shown in graph 900, where the Kalman filter algorithm enables precise position determination even through double integration of acceleration measurements.

[0094] In some examples, the Kalman filter algorithm may continuously update position estimates based on new sensor data from both image sensor 173 and the IMU sensor 712, providing real-time three-dimensional positioning feedback for material processing control. Processing circuitry 182 may reset state vectors in the Kalman filter algorithm when electrical contact is detected between torch 106 and workpiece 312, providing reference points that enhance position tracking accuracy and reduce cumulative drift effects during extended material processing operations.

[0095] Graph 900 may demonstrate that acceleration sensor measurements alone may track torch 106 tip position relatively accurately for short periods of time through the Kalman filter processing, even after double integration of the acceleration data. The position accuracy achieved through the Kalman filter algorithm may be suitable for material processing applications where position tracking resolution of approximately 1 mm may be adequate for operational control and safety monitoring during material processing operations.

[0096] FIG. 10 is a graph 1000 illustrating the high angular resolution capabilities of the IMU sensor 712 for detecting torch 106 orientation changes during material processing operations. Graph 1000 include roll Euler angles plotted over time, with measurements extending over approximately 30 seconds to illustrate the IMU sensor 712 capability to track discrete angular changes with precision. The data indicated angular step changes of approximately 0.2 degrees, demonstrating the resolution capability of the IMU sensor 712 to detect small orientation changes of torch 106 during material processing operations. The high angular resolution capability may be valuable for determining focus distance variations that occur when the torch orientation 214 changes relative to workpiece 312, where the focus distance 218 may shift according to the relationship Az = a sin(0), where a represents a constant and 0 represents the torch orientation angle measured by the IMU sensor 712.

[0097] Processing circuitry 182 may utilize the angular measurements demonstrated in graph 1000 to calculate focus distance adjustments needed to compensate for torch 106 orientation changes during material processing operations. The capability to detect 0.2 degree angular steps may enable processing circuitry 182 to determine when the torch orientation 214 has changed sufficiently to require focus distance corrections, even when electrical contact between torch 106 and workpiece 312 is maintained.

[0098] The sensor fusion capabilities may incorporate the angular step detection demonstrated in graph 1000 with triangulation-based distance measurements from imageAttorney Docket No. 5117.0025W01 sensor 173 to provide comprehensive three-dimensional positioning information. The combination of high-resolution angular measurements from the IMU sensor 712 and distance measurements from triangulation calculations may enable processing circuitry 182 to determine both the distance between torch 106 and workpiece 312 and the angular orientation of torch 106 relative to workpiece 312.

[0099] Controller 180 may be configured to automatically adjust at least one operational parameter of torch 106 based on the angular orientation changes detected through the angular step resolution capability demonstrated in graph 1000. The operational parameter adjustments may include focus distance corrections, laser power modifications, and process gas flow adjustments that compensate for the geometric changes that occur when the torch orientation 214 varies during material processing operations.

[0100] FIGS. 11 and 12 are respective graphs 1100 and 1200 illustrating spectrometer output versus wavelength at 200W and 500W laser power, respectively. Graphs 1100 and 1200 demonstrate the spectral analysis capabilities of spectrometer 175 during material processing operations on stainless steel workpieces. The graphs show signal strength from approximately 1000 to 9000 units plotted against wavelength from 200 to 1100 nanometers (nm) with a prominent peak at approximately 960.18 nm wavelength and a broad elevation between 600-800 nm wavelengths at both power levels. The peak at 960.18 nm represents a characteristic spectral emission line corresponding to elemental components in the stainless steel workpiece. Processing circuitry 182 may identify this spectral signature for material identification through laser-induced breakdown spectroscopy (LIBS) analysis.

[0101] Spectrometer 175 may perform LIBS to identify workpiece material composition based on spectral signatures generated during laser processing operations. The LIBS technique utilizes high-energy laser pulses from torch 106 to create plasma from the workpiece material, where atoms emit light at specific wavelengths corresponding to the elemental composition. Spectrometer 175 analyzes the emitted light to determine material composition through characteristic spectral emission lines.

[0102] The increased laser power at 500W compared to 200W enhances signal strength and spectral resolution. Higher laser power creates more intense plasma formation, resulting in stronger spectral emission lines that enable more accurate material identification. The enhanced signal strength improves the signal-to-noise ratio, enabling processing circuitry 182 to detect spectral features with greater precision.

[0103] The spectral signature characteristics enable processing circuitry 182 to distinguish stainless steel workpieces from other materials based on the unique combinationAttorney Docket No. 5117.0025W01 of the peak at 960.18 nm and the broad elevation between 600-800 nm wavelengths. This material identification capability enables controller 180 to automatically adjust operational parameters of torch 106 based on detected workpiece composition, optimizing laser power, process gas flow, and other processing parameters.

[0104] Controller 180 may utilize the material identification capabilities provided by spectrometer 175 across different power levels to enhance material processing quality and safety. The spectral analysis results enable controller 180 to verify correct material processing and adjust operational parameters accordingly. The material identification capability also enables detection of workpiece contamination, coatings, or unexpected material compositions that could affect processing quality or generate hazardous emissions.

[0105] Processing circuitry 182 may compare spectral signatures obtained at different laser power levels to provide comprehensive material analysis during material processing operations. The comparison between spectral patterns at 200W and 500W laser power levels enables verification of material identification accuracy and detection of changes in workpiece composition during processing operations.

[0106] FIG. 13 is a graph 1300 illustrating spectrometer output versus wavelength at 800W laser power on an aluminum coupon. Graph 1300 shows signal strength from approximately 1000 to 1800 units plotted against wavelength from 200 to 1100 nanometers (nm) with peaks distributed across different wavelength regions, creating a spectral pattern that differs completely from the stainless steel spectral signatures shown in FIGS. 11 and 12. The distinct peak distribution enables processing circuitry 182 to identify aluminum workpieces through their characteristic spectral emission lines corresponding to aluminum elemental composition.

[0107] Spectrometer 175 may perform LIBS to identify aluminum workpiece material composition based on spectral signatures generated during laser processing operations. The aluminum spectral signature exhibits peak positions and intensity distributions distinctly different from the stainless steel characteristics, enabling processing circuitry 182 to distinguish between aluminum and stainless steel workpieces for automatic material identification.

[0108] The 800W laser power level provides enhanced plasma formation intensity that enables detection of multiple elemental components within aluminum workpiece materials. Higher laser power creates more intense plasma conditions generating stronger spectral emission lines from both primary aluminum components and trace alloying elements, providing comprehensive material composition analysis.Attorney Docket No. 5117.0025W01

[0109] Controller 180 may utilize the material identification capabilities provided by spectrometer 175 at high power levels to enhance material processing quality and safety. The spectral analysis results enable controller 180 to verify correct aluminum material processing and adjust operational parameters accordingly. The material identification capability also enables detection of workpiece coatings or contamination that could affect processing quality or generate hazardous emissions.

[0110] Sensor suite 196 may include a volatile organic compound (VOC) sensor for detecting hazardous gas emissions during material processing operations on aluminum workpieces or other materials that may have coatings or organic contamination. The VOC sensor may be configured to detect organic vapors, smoke, or hazardous chemical substances that may be generated when high-intensity laser energy contacts coated materials, painted surfaces, or workpieces with organic contamination. The VOC sensor may provide safety monitoring capabilities that complement the material identification provided by spectrometer 175 through LIBS analysis.

[0111] Spectrometer 175 may be configured to detect coating or organic materials that may generate VOC or hazardous chemical substances during laser processing operations. When torch 106 contacts workpieces with coatings, paints, adhesives, or other organic materials, the high-intensity laser energy may vaporize these materials and create spectral emissions that differ from the base workpiece material. Processing circuitry 182 may be configured to identify spectral signatures that indicate the presence of organic materials or coatings based on characteristic emission lines that correspond to carbon, hydrogen, oxygen, or other elements commonly found in organic compounds.

[0112] Controller 180 may be configured to automatically adjust operational parameters or provide safety warnings when spectrometer 175 detects spectral signatures indicating the presence of organic materials or coatings that may generate hazardous emissions. The spectral analysis results may enable controller 180 to identify when aluminum workpieces include surface treatments, coatings, or contamination that could affect processing quality or generate safety concerns during material processing operations.

[0113] The VOC sensor within sensor suite 196 may work in conjunction with spectrometer 175 to provide comprehensive monitoring of both material composition and hazardous gas emissions during material processing operations. The VOC sensor may detect organic vapors or hazardous chemical substances that are released when laser energy contacts coated or contaminated surfaces, while spectrometer 175 may identify the spectral signatures that indicate the presence of organic materials through LIBS analysis. The combination ofAttorney Docket No. 5117.0025W01VOC detection and spectral analysis may provide comprehensive safety monitoring that protects operators from exposure to hazardous emissions during material processing operations.

[0114] Processing circuitry 182 may be configured to correlate spectral signatures from spectrometer 175 with VOC sensor readings to provide enhanced material safety assessment during material processing operations. When spectrometer 175 detects spectral signatures indicating organic materials or coatings, processing circuitry 182 may monitor VOC sensor readings to determine if hazardous gas emissions are being generated. The correlation between spectral analysis and VOC detection may enable controller 180 to make informed decisions about continuing material processing operations or implementing safety measures to protect operators from hazardous exposure.

[0115] The material identification capabilities demonstrated through the aluminum spectral signatures may enable controller 180 to optimize material processing parameters specifically for aluminum workpieces. The unique spectral pattern characteristics of aluminum may enable controller 180 to adjust laser power, process gas flow, wobble frequency, and other operational parameters to values that are optimized for aluminum material processing. The automatic parameter optimization based on material identification may improve processing quality and efficiency while reducing the risk of processing defects or safety concerns during aluminum material processing operations.

[0116] FIG. 14 illustrates an example experimental apparatus 1400 used to demonstrate spectral analysis capabilities of spectrometer 175 during material processing operations, providing controlled testing environment for laser- induced breakdown spectroscopy (LIBS) measurements and material composition determination. Torch 1402 generates laser energy at select power levels (e.g., from 200W to 800W) to create plasma formation on workpiece surfaces, enabling spectral emission generation for material identification. Torch mount 1408 provides stable positioning and precise control of torch 1402 position and orientation relative to target workpieces. Spectrometer 1404 captures and analyzes spectral emissions across wavelength ranges from 200 nm to 1100 nm, providing comprehensive spectral coverage for LIBS analysis and real-time material composition determination. Fiber optic probe 1410 couples between spectrometer 1404 and plasma generation region to collect spectral emissions while maintaining safe positioning away from high-intensity laser processing area. Optical axis 1412 defines primary light collection path along which fiber optic probe 1410 captures spectral emissions and transmits to spectrometer 1404, positioned to optimizeAttorney Docket No. 5117.0025W01 spectral emission collection while avoiding interference with laser energy delivery from torch 1402.

[0117] Processing circuitry 182 analyzes spectral data collected through fiber optic probe 1410 to determine material composition during welding operations, enabling identification of workpiece materials such as stainless steel or aluminum based on characteristic emission lines at specific wavelengths. Controller 180 automatically adjusts operational parameters of torch 1402 based on detected workpiece composition.

[0118] Fiber optic probe 1410 collects spectral emissions corresponding to elemental components in workpiece materials, where plasma created by torch 1402 causes atoms to emit light at wavelengths corresponding to their elemental composition. Spectrometer 1404 detects spectral signatures indicating welding quality, plasma temperature, elemental composition, and process stability that correlate with welding penetration depth and fusion quality.

[0119] Experimental apparatus 1400 enables validation of spectral analysis accuracy across different laser power levels and workpiece materials through controlled positioning provided by torch mount 1408. Optical axis 1412 positioning optimizes plasma emission capture while minimizing background light interference and avoiding reflected laser energy that could compromise spectral measurement accuracy.

[0120] Controller 180 utilizes spectral analysis results to develop material- specific processing parameters that optimize welding quality for different workpiece compositions, creating material identification databases that correlate specific spectral patterns with optimal operational parameters for enhanced welding quality and process efficiency.

[0121] FIG. 15 is a graph 1500 illustrating an experimental requits of a confusion matrix for twelve power / gap combinations may demonstrate the machine learning classification capabilities of modular handheld laser system 100 for automatic parameter identification during material processing operations. Graph 1500 shows the relationship between predicted labels on the x-axis and true labels on the y-axis for power and gap distance combinations ranging from 200W-0mm to 500W-2mm, with darker squares indicating higher numbers of correct predictions across various power and gap distance combinations. Graph 1500 validates the capability of machine learning algorithms to identify operational parameters based on sensor data collected during material processing operations.

[0122] Processing circuitry 182 was configured to implement machine learning algorithms including a LightGBM classifier model for process parameter prediction during material processing operations. The LightGBM classifier model was trained using acousticAttorney Docket No. 5117.0025W01 sensor data collected by acoustic sensor 167 during material processing operations at various power levels and gap distances. The machine learning capabilities enabled processing circuitry 182 to automatically identify operational parameters based on acoustic signatures detected during material processing operations, providing feedback for automatic parameter adjustment and quality control.

[0123] Acoustic sensor 167 operates in ultrasonic frequency range from 100 Hz to 80 kHz for enhanced material processing monitoring. The extended frequency range of acoustic sensor 167 captures acoustic signatures that extend beyond human hearing range, providing comprehensive acoustic data that correlates with specific power and gap distance combinations during material processing operations. The ultrasonic frequency range capability enabled acoustic sensor 167 to detect acoustic patterns that are not detectable by conventional audio sensors, providing enhanced data for machine learning algorithm training and parameter identification.

[0124] The LightGBM classifier model achieved 100% accuracy per file for gap distance identification based on acoustic sensor data collected. The gap distance classification accuracy demonstrated in graph 1500 may enable processing circuitry 182 to determine the distance between torch 106 and workpiece 312 based on acoustic signatures alone, providing an alternative distance measurement method that complements triangulation-based measurements from image sensor 173. The 100% gap distance accuracy validates the effectiveness of acoustic signature analysis for proximity detection during material processing operations.

[0125] The LightGBM classifier model achieved 75% accuracy for power classification across the power and gap distance combinations shown in graph 1500. The power classification accuracy enables processing circuitry 182 to identify the actual laser power being delivered during material processing operations based on acoustic signatures generated during plasma formation and material interaction. The power identification capability may provide feedback to controller 180 for verifying that intended power settings are being achieved and for detecting variations in laser output that may affect material processing quality.

[0126] In some examples, controller 180 may be configured to automatically adjust multiple operational parameters including laser power, process gas flow rate, and wobble frequency based on the determined distance and parameter identification results from the machine learning algorithms. The automatic parameter adjustment capabilities may utilize the gap distance and power classification results from the LightGBM classifier model to optimizeAttorney Docket No. 5117.0025W01 material processing conditions in real-time. Controller 180 may adjust laser power output to maintain consistent processing conditions when gap distance variations are detected through acoustic signature analysis, ensuring optimal energy delivery to workpiece 312 regardless of torch 106 positioning variations.

[0127] Controller 180 may automatically adjust process gas flow rate based on the power and gap distance combinations identified through the machine learning classification results shown in graph 1500. The process gas flow rate adjustments may optimize shielding gas coverage and plasma stability based on the detected operational parameters, ensuring consistent material processing quality across different power and distance conditions. The automatic process gas flow rate adjustment may compensate for changes in plasma characteristics that occur when power levels or gap distances vary during material processing operations.

[0128] Controller 180 may automatically adjust wobble frequency based on the machine learning classification results to optimize material processing characteristics for the detected power and gap distance combinations. The wobble frequency adjustments may modify the laser beam oscillation patterns to maintain consistent weld penetration and fusion characteristics when operational parameters vary during material processing operations. The automatic wobble frequency adjustment may ensure that material processing quality remains consistent even when torch 106 positioning or power output varies during handheld material processing operations.

[0129] A method of controlling torch 106 during material processing may include automatically adjusting multiple operational parameters including laser power, process gas flow rate, and wobble frequency based on the determined distance and machine learning classification results. The method may utilize acoustic signatures detected by acoustic sensor 167 operating in the ultrasonic frequency range from 100 Hz to 80 kHz to provide input data for the LightGBM classifier model. The method may enable real-time parameter optimization based on acoustic signature analysis that identifies both gap distance and power level conditions during material processing operations.

[0130] Processing circuitry 182 may be configured to continuously update the LightGBM classifier model based on new acoustic sensor data collected during material processing operations, enabling adaptive learning that improves classification accuracy over time. The continuous learning capability may enable the machine learning algorithms to adapt to different workpiece materials, environmental conditions, or operational variations that may affect acoustic signature characteristics during material processing operations. The adaptiveAttorney Docket No. 5117.0025W01 learning may enhance the reliability of automatic parameter adjustment by improving the accuracy of gap distance and power classification based on accumulated operational experience.

[0131] The machine learning classification capabilities demonstrated in graph 1500 may complement other sensor measurements within sensor suite 196 by providing acoustic-based parameter identification that operates independently of optical or electromagnetic sensing methods. The acoustic signature analysis may provide parameter identification capabilities when other sensors may be compromised by environmental conditions, workpiece surface characteristics, or geometric constraints that affect optical triangulation or capacitance-based measurements. The combination of machine learning classification with other sensor measurements may provide comprehensive parameter monitoring and automatic adjustment capabilities that enhance material processing quality and operational efficiency.

[0132] FIG. 16 is a photo 1600 illustrating the cross-sectional characteristics of weld formations created during material processing operations with modular handheld laser system 100. Photo 1600 shows a weld cross-section having a total depth 1608 that illustrates the varying penetration depth characteristics achieved through selected laser processing parameters and wobble patterns.

[0133] A central penetration depth 1602 may be visible in the middle region of the weld cross-section shown in photo 1600. The central penetration depth 1602 may represent the maximum depth of material fusion achieved along the centerline of the laser beam path during material processing operations. The central penetration depth 1602 may correspond to the region where laser energy density is highest and where the primary laser beam creates the deepest material penetration during welding operations.

[0134] A wobble penetration depth 1604 may be visible at the outer regions of the weld cross-section shown in photo 1600. The wobble penetration depth 1604 may represent the depth of material fusion achieved in regions where the laser beam oscillates during wobble pattern operations. The wobble penetration depth 1604 may differ from the central penetration depth 1602 due to the distributed energy characteristics that occur when the laser beam follows oscillating patterns during material processing operations.

[0135] A width 1606 may be indicated at the bottom of the weld cross-section shown in photo 1600. The width 1606 may represent the overall lateral extent of the weld formation, encompassing both the central penetration region and the wobble penetration regions. The width 1606 may be determined by the wobble amplitude and pattern characteristics selectedAttorney Docket No. 5117.0025W01 during material processing operations, where larger wobble amplitudes may create wider weld formations.

[0136] Photo 1600 may demonstrate different penetration depths achieved during the welding process based on the spatial distribution of laser energy during material processing operations. The central penetration depth 1602 may be greater than the wobble penetration depth 1604 due to the concentrated energy delivery along the primary laser beam path, while the wobble penetration depth 1604 may provide broader fusion coverage across the width 1606 of the weld formation. The penetration depth variations shown in photo 1600 may illustrate how wobble patterns affect energy distribution and material fusion characteristics.

[0137] The measurements demonstrated in photo 1600 may serve as ground truth data for machine learning regression models that predict weld depth from sensor measurements collected during material processing operations. The central penetration depth 1602 and the wobble penetration depth 1604 measurements may provide reference values that enable processing circuitry 182 to correlate sensor data from sensor suite 196 with actual weld penetration characteristics achieved during material processing operations. The ground truth data may enable training of machine learning algorithms that predict weld quality and penetration depth based on acoustic signatures detected by acoustic sensor 167, spectral emissions analyzed by spectrometer 175, and other sensor measurements collected during welding operations.

[0138] Processing circuitry 182 may be configured to utilize machine learning regression models that correlate sensor measurements with the central penetration depth 1602 and the wobble penetration depth 1604 characteristics demonstrated in photo 1600. The machine learning regression models may be trained using acoustic sensor data from acoustic sensor 167 operating in the ultrasonic frequency range from 100 Hz to 80 kHz, where acoustic signatures may correlate with the penetration depth characteristics achieved during material processing operations. The regression models may enable processing circuitry 182 to predict both the central penetration depth 1602 and the wobble penetration depth 1604 based on realtime sensor measurements during welding operations.

[0139] Controller 180 may be configured to automatically adjust operational parameters of torch 106 based on predicted penetration depth values calculated using the machine learning regression models trained with ground truth data from measurements such as those shown in photo 1600. The automatic parameter adjustments may include laser power modifications, wobble frequency changes, and wobble amplitude adjustments that optimize the central penetration depth 1602 and the wobble penetration depth 1604 to achieve desiredAttorney Docket No. 5117.0025W01 weld characteristics. Controller 180 may adjust laser power output to increase or decrease the central penetration depth 1602 based on the predicted depth values calculated from sensor measurements during material processing operations.

[0140] The width 1606 measurements demonstrated in photo 1600 may provide additional ground truth data for machine learning models that predict weld geometry characteristics based on sensor measurements. Processing circuitry 182 may be configured to correlate the width 1606 measurements with wobble amplitude and frequency parameters used during material processing operations, enabling prediction of weld width characteristics based on operational parameter settings. The width 1606 prediction capability may enable controller 180 to adjust wobble parameters to achieve desired weld width characteristics during material processing operations.

[0141] The machine learning regression models may achieve average error of approximately 0.16 mm when predicting penetration depth based on sensor measurements correlated with ground truth data such as the measurements shown in photo 1600. The prediction accuracy may enable approximately 46% of predictions to be within 0.1 mm of correct depth values, with approximately 70% of predictions within 0.2 mm of correct depth values. The high prediction accuracy may enable processing circuitry 182 to provide reliable weld depth estimation during material processing operations based on real-time sensor measurements from sensor suite 196.

[0142] The ground truth data provided by measurements such as the central penetration depth 1602 and the wobble penetration depth 1604 shown in photo 1600 may enable processing circuitry 182 to distinguish between different penetration characteristics achieved through various welding parameters. The central penetration depth 1602 measurements may provide reference data for training machine learning models to predict maximum penetration depth, while the wobble penetration depth 1604 measurements may provide reference data for predicting distributed penetration characteristics across the width 1606 of the weld formation. The comprehensive ground truth data may enable machine learning models to predict both localized and distributed weld characteristics based on sensor measurements collected during material processing operations.

[0143] FIG. 17 is a graph 1700 illustrating a predicted vs true depth. Graph 1700 demonstrates the accuracy of machine learning regression models for predicting central penetration depth during material processing operations with modular handheld laser system 100. Graph 1700 shows predicted depth values plotted on the vertical axis against true depth values plotted on the horizontal axis, and indicates the ideal one-to-one correlation betweenAttorney Docket No. 5117.0025W01 predicted and true depth values, representing perfect prediction accuracy where predicted values exactly match measured values. Data points may be scattered around the diagonal line in graph 1700, showing the actual relationship between predicted and measured central penetration depth values achieved during material processing operations. The distribution of data points relative to the diagonal line indicates the accuracy and precision of the machine learning regression model for predicting weld penetration characteristics.

[0144] Processing circuitry 182 may be configured to implement LightGBM regression models that predict the central penetration depth 1602 with average error of approximately 0.16 mm based on acoustic sensor data collected by acoustic sensor 167 during material processing operations. The LightGBM regression model may utilize acoustic signatures detected in the ultrasonic frequency range from 100 Hz to 80 kHz to correlate sound patterns with actual penetration depth measurements, enabling real-time prediction of weld depth characteristics during material processing operations. The average error of approximately 0.16 mm demonstrated in graph 1700 validate the effectiveness of acoustic signature analysis for weld quality assessment.

[0145] The LightGBM regression model may achieve approximately 88% of predictions within 0.3 mm of correct depth for central penetration depth estimation, as demonstrated by the data point distribution in graph 1700. The high prediction accuracy may enable processing circuitry 182 to provide reliable weld depth estimation during material processing operations based on real-time acoustic sensor measurements from acoustic sensor 167. The 88% accuracy within 0.3 mm tolerance may demonstrate the machine learning capability for real-time weld quality assessment that enables automatic process control and quality monitoring during material processing operations.

[0146] Controller 180 may be configured to automatically adjust at least one operational parameter of torch 106 based on the central penetration depth predictions calculated using the LightGBM regression model. The at least one operational parameter may include laser power output adjustments that optimize penetration depth characteristics based on the predicted depth values shown in graph 1700. Controller 180 may increase laser power output when the predicted central penetration depth falls below desired target values, or decrease laser power output when the predicted penetration depth exceeds target specifications, enabling automatic control of weld depth characteristics during material processing operations.

[0147] The laser power output adjustments may be implemented in real-time based on the acoustic signatures detected by acoustic sensor 167 and processed through the LightGBM regression model to predict central penetration depth values. Controller 180 may utilize theAttorney Docket No. 5117.0025W01 predicted depth values to maintain consistent weld penetration characteristics even when workpiece conditions, torch positioning, or other operational variables change during material processing operations. The automatic laser power output adjustment capability may ensure optimal energy delivery to achieve desired penetration depth targets while preventing excessive penetration that could compromise weld quality or workpiece integrity.

[0148] The machine learning capability for real-time weld quality assessment demonstrated in graph 1700 enable processing circuitry 182 to provide continuous monitoring of weld penetration characteristics without requiring destructive testing or postprocess inspection. The real-time prediction capability may allow operators to receive immediate feedback on weld quality during material processing operations, enabling corrective actions to be taken before weld defects occur. The non-destructive weld quality assessment provided by the acoustic signature analysis and machine learning prediction may enhance material processing efficiency by reducing the need for post-process inspection and rework operations.

[0149] Graph 1700 demonstrates the reliability of acoustic-based penetration depth prediction across a range of weld depth values from approximately 0.4 to 1.2 millimeters. The consistent prediction accuracy across the depth range may enable the LightGBM regression model to provide reliable weld quality assessment for various material thicknesses and processing conditions encountered during material processing operations. The broad depth range coverage may ensure that the machine learning prediction capability remains effective across different welding applications and workpiece configurations.

[0150] Processing circuitry 182 may be configured to continuously update the LightGBM regression model based on new acoustic sensor data and corresponding penetration depth measurements collected during material processing operations. The continuous learning capability may enable the machine learning algorithms to adapt to different workpiece materials, environmental conditions, or operational variations that may affect the relationship between acoustic signatures and penetration depth characteristics. The adaptive learning may enhance the accuracy of penetration depth prediction over time by incorporating operational experience and improving the correlation between acoustic patterns and actual weld depth measurements.

[0151] FIG. 18 illustrates a graph 1800 of predicted vs true maximum depth. Graph 1800 demonstrates the accuracy of machine learning regression models for predicting maximum penetration depth during wobble pattern welding operations, similar to the central depth predictions shown in FIG. 17. The graph 1800 shows predicted maximum depth valuesAttorney Docket No. 5117.0025W01 plotted against true maximum depth values ranging from approximately 0.4 to 1.6 millimeters, with data points scattered around a diagonal line indicating ideal one-to-one correlation between predicted and measured maximum penetration depths.

[0152] The processing circuitry 182 may be configured to implement LightGBM regression models that predict the wobble penetration depth 1604 with approximately 80% of predictions within 0.3 mm of correct depth, utilizing acoustic signatures detected by the acoustic sensor 167 in the ultrasonic frequency range from 100 Hz to 80 kHz. This accuracy level, while slightly lower than the central penetration depth predictions demonstrated in FIG. 17, validates the effectiveness of acoustic signature analysis for assessing wobble pattern weld quality across the extended depth range.

[0153] The controller 180 may be configured to automatically adjust operational parameters including laser power output, process gas flow rate, and wobble frequency based on the maximum penetration depth predictions demonstrated in the graph 1800. These automatic adjustments may optimize wobble pattern effectiveness and ensure consistent weld quality, providing real-time feedback similar to the central depth control capabilities while accommodating the broader depth range requirements of wobble pattern operations.

[0154] The processing circuitry 182 may be configured to correlate the wobble maximum penetration depth predictions with the central penetration depth 1602 predictions to provide comprehensive weld geometry assessment. This correlation capability, building upon the central depth prediction accuracy shown in FIG. 17, may enable the controller 180 to optimize wobble pattern parameters for desired penetration depth profiles that balance maximum depth achievement with consistent fusion characteristics across the weld formation width.

[0155] FIG. 19 is a graph 1900 illustrating the relative contributions of different frequency ranges to machine learning model accuracy during material processing operations with the modular handheld laser system 100. Graph 1900 demonstrates how acoustic signatures detected across different frequency ranges contribute to the effectiveness of machine learning algorithms for weld quality prediction and operational parameter identification. For example, acoustic sensor data analysis reveals approximately 28.79% contribution from human hearing range frequencies while approximately 71.21% contribution comes from frequencies outside the human hearing range. This frequency contribution analysis validates the importance of ultrasonic frequency sensing capabilities provided by the acoustic sensor 167 operating in the frequency range from 100 Hz to 80 kHz during material processing operations.Attorney Docket No. 5117.0025W01

[0156] The processing circuitry 182 may be configured to analyze acoustic signatures across the full frequency range from 100 Hz to 80 kHz, where graph 1900 demonstrates that frequencies outside the human hearing range contribute significantly more information for machine learning model accuracy. The ultrasonic frequency sensing may detect acoustic patterns that correlate with material processing characteristics such as penetration depth, weld quality, and operational parameters that are not detectable through conventional audio frequency monitoring.

[0157] The machine learning model accuracy may drop significantly when restricted to human hearing frequency range, as indicated by the frequency contribution analysis shown by graph 1900. The 71.21% contribution from frequencies outside the human hearing range demonstrates that ultrasonic acoustic signatures provide information that is not available through audible frequency monitoring alone.

[0158] The controller 180 may be configured to utilize the comprehensive frequency range analysis demonstrated by graph 1900 to optimize acoustic signature processing for enhanced material processing control. The frequency contribution information may enable the controller 180 to prioritize ultrasonic frequency data processing when implementing automatic parameter adjustments based on acoustic signature analysis, including laser power output, process gas flow rate, and wobble frequency adjustments.

[0159] The comprehensive frequency range analysis demonstrated by graph 1900 validates the design decision to incorporate the acoustic sensor 167 with extended frequency range capabilities rather than conventional audio sensors limited to human hearing range. The 71.21% contribution from ultrasonic frequencies enables machine learning models to achieve higher accuracy for gap distance identification, power classification, and penetration depth prediction compared to models utilizing only audible frequency acoustic signatures.

[0160] US Patent Application Publication No. US-20230121254-A1, titled “Laser Processing System with Integrated Sensors,” discloses a laser welding device and a laser welding method in which when an unacceptable welding defect in laser welding is detected, measures for quickly and appropriately maintaining a welding strength of a welding portion where the welding defect is detected are taken. The entirety of US-20230121254-Al is hereby incorporated by reference herein for all purposes.

[0161] A number of implementations have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the disclosure. Accordingly, other implementations are within the scope of the following claims.

Claims

Attorney Docket No. 5117.0025W01CLAIMS1. A handheld laser tool system, comprising: a handheld laser tool having a body portion with a laser beam pathway and a forward laser beam exit; a sensor suite operatively coupled to the handheld laser tool, the sensor suite including an image sensor configured to detect laser light reflected from a workpiece and a processing circuit configured to determine a distance between the handheld laser tool and the workpiece based on triangulation using the laser light; and a controller operatively coupled to the sensor suite and configured to adjust at least one operational parameter of the handheld laser tool based on the distance.

2. The handheld laser tool system of claim 1, wherein the sensor suite further comprises an inertial measurement unit (IMU) configured to detect orientation and movement of the handheld laser tool.

3. The handheld laser tool system of claim 2, wherein the processing circuit is configured to perform sensor fusion between data from the image sensor and data from the IMU to determine a three-dimensional position of the handheld laser tool relative to the workpiece.

4. The handheld laser tool system of claim 1, wherein the image sensor is configured to detect laser light from a marker laser beam projected onto the workpiece.

5. The handheld laser tool system of claim 4, wherein the marker laser beam has a wavelength different from a primary laser beam used for material processing.

6. The handheld laser tool system of claim 1, wherein the triangulation is based on a mutual angle between a laser vector and an optical axis of the image sensor.

7. The handheld laser tool system of claim 6, wherein the mutual angle is between 10 degrees and 45 degrees.

8. The handheld laser tool system of claim 1, wherein the processing circuit is configured to determine the distance with a resolution of 200 micrometers or better.Attorney Docket No. 5117.0025W019. The handheld laser tool system of claim 1, wherein the at least one operational parameter comprises laser power output.

10. The handheld laser tool system of claim 1, wherein the sensor suite further comprises an acoustic sensor configured to detect sound generated during material processing operations.

11. The handheld laser tool system of claim 10, wherein the processing circuit is configured to correlate acoustic signatures from the acoustic sensor with the distance to assess material processing quality.

12. The handheld laser tool system of claims 1 or 10, wherein the controller is configured to automatically adjust multiple operational parameters including laser power, process gas flow rate, and wobble frequency based on the distance.

13. The handheld laser tool system of claim 3, wherein the sensor fusion utilizes a Kalman filter algorithm to combine data from the image sensor and the IMU.

14. A method of controlling a handheld laser tool during material processing, comprising: directing laser light from the handheld laser tool toward a workpiece; detecting, with an image sensor, laser light reflected from the workpiece; determining, with a processing circuit, a distance between the handheld laser tool and the workpiece based on triangulation using the laser light; and automatically adjusting at least one operational parameter of the handheld laser tool based on the distance.

15. The method of claim 14, further comprising detecting, with an inertial measurement unit (IMU), orientation and movement of the handheld laser tool.

16. The method of claim 15, further comprising performing sensor fusion between data from the image sensor and data from the IMU to determine a three-dimensional position of the handheld laser tool relative to the workpiece.Attorney Docket No. 5117.0025W0117. The method of claim 14, wherein the laser light comprises a marker laser beam having a wavelength different from a primary laser beam used for material processing.

18. The method of claim 16, wherein the sensor fusion utilizes a Kalman filter algorithm to combine data from the image sensor and the IMU.

19. The method of claims 14 or 17, wherein automatically adjusting the at least one operational parameter comprises adjusting multiple operational parameters including laser power, process gas flow rate, and wobble frequency based on the distance.

20. A handheld laser tool system, comprising: a handheld laser tool configured to generate laser pulses for creating plasma from a workpiece material; a spectrometer operatively coupled to the handheld laser tool and configured to analyze light emitted from the plasma to determine elemental composition of the workpiece material through laser-induced breakdown spectroscopy; and a processing circuit configured to identify a material composition based on characteristic spectral emission lines detected by the spectrometer.

21. The handheld laser tool system of claim 20, wherein the spectrometer is configured to analyze spectral emissions across wavelength ranges from 200 nm to 1100 nm.

22. The handheld laser tool system of claim 20, further comprising a fiber optic probe configured to collect spectral emissions from the plasma and transmit the spectral emissions to the spectrometer.

23. The handheld laser tool system of claim 20, wherein the processing circuit is configured to distinguish between different workpiece materials based on unique spectral signature patterns.

24. The handheld laser tool system of claim 20, further comprising a controller configured to automatically adjust operational parameters of the handheld laser tool based on the material composition.Attorney Docket No. 5117.0025W0125. The handheld laser tool system of any one of claims 20 through 24, wherein the spectrometer is configured to detect spectral signatures indicating presence of organic materials or coatings that may generate hazardous emissions during laser processing operations.

Images