Test and measurement devices, and methods of using ai assistant in test and measurement devices

Abstract

To shorten the test time using AI.SOLUTION: A test and measurement device 10 includes one or more ports 28 to connect to a device under test (DUT) 12, a user interface 22 having one or more control devices, a display 24, a storage 26, one or more processors 20. The processors 20 receive test signals from the DUT 12 through the one or more ports 28 during test of the DUT 12, use the test signals to generate test data, display test data on the display 24, display on the user interface 22 a control button indicating that an artificial intelligence (AI) assistant is available, receive an input through the control button to start the AI assistant, provide, on the user interface 22, regions allowing the user to interact with the AI assistant, and apply a machine learning model represented by the AI assistant to provide the user with additional information related to one or more of the test and the DUT 12.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to artificial intelligence (AI), and more particularly to an AI assistant for use in a test measurement system to enhance insight, improve efficiency, and enable predictive maintenance.

Background Art

[0002] AI and machine learning (ML) have become powerful tools for data analysis and extracting insights from test data. AI-driven models can autonomously interpret complex data patterns, enabling more efficient and accurate analysis of test results. Machine learning has the ability to adapt and improve performance over time, playing an important role in predictive maintenance by identifying potential device failures before they occur.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] Currently, ML and AI can make these predictions and perform some of this type of data analysis, but it can currently take a long time for testing. Furthermore, in the current approach, pre-trained models are used. To use a pre-trained model, it is necessary to collect a large amount of learning data, train the model, and then perform verification. This process takes more time than expected and affects the user's workflow in order to accommodate the test cycle.

Means for Solving the Problems

[0005] Embodiments of the present application include an AI-enabled test measurement system that improves the functionality of the test measurement system, provides unparalleled insights, and improves testing and cost efficiency. Embodiments of the present application can autonomously interpret complex data patterns, enabling more accurate analysis of test results. Embodiments of the present application improve performance over time, play an important role in predictive maintenance, and can identify potential device failures before they occur. The embodiments employ a machine learning model that grows and learns in real time while the user continues to use the test measurement device during design and verification.

[0006] Embodiments of the present application include an "AI assistant", which refers to a machine learning model. In the context of the present application, these terms are used interchangeably, so references to the AI assistant refer to the user interface of the machine learning model. Unlike previous machine learning models, embodiments of the present application do not change the user's workflow. When the user runs tests on a device under test (DUT), the embodiments collect data from the user's usage of the test measurement device, first define and create a model, and then train the model. This embodiment provides a model that can be deployed across multiple test measurement endpoints, enabling consistent results across multiple devices. Using version management, other endpoints can update the machine learning model without overwriting previous versions. Similarly, the embodiments extend the model to operate across multiple measurements, leveraging the same captured data to provide insights. For a system that includes all endpoints and model storage, external users can use a subscription service, enabling these external users to access the machine learning model according to their needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Embodiments for Carrying Out the Invention

[0008] FIG. 1 shows the architecture of an embodiment of a system for creating and using an embodiment of an AI assistant. The test measurement device 10 is specified as an oscilloscope or a "scope" in the present application, but may be composed of any test measurement device. It is not limited to an oscilloscope. In a test environment, the oscilloscope 10 is connected to the DUT 12. This oscilloscope may have one or more processors (such as 20), a user interface (U / I) 22 having one or more operating devices, a display unit (display) 24, and a storage 26. The operating device may include, for example, buttons, knobs, and sliders, or may exist on the device body as a touch "button" on the display as part of the user interface, or may exist as both. The display unit (display) may include at least a part of the user interface. When referring to storage, it may include both local storage and links to external storage, including one or more databases (such as 14) or cloud storage 16. Cloud storage may be connected to various elements of the architecture, but is likely to be connected to at least the database 14 and the neural network 18.

[0009] When the user operates the test measurement device 10, the test measurement device will receive a test signal from the DUT 12. In some cases, one or more processors 20 apply a signal to the DUT 12 to start the test. The DUT 12 generates a test signal in response to the applied signal. In another case, the test measurement device may receive a test signal from the DUT without the need to apply a signal to the DUT to start the test. The test signal received by the test measurement device may be composed of an analog signal or a digital signal received via the port 28. As a general example, the DUT 12 generates analog signals, and the test measurement device 10 converts them into test data such as waveforms. Then, the test measurement device displays the test data on the display unit of the device.

[0010] During the use of the test and measurement apparatus as described above, as an example, the test and measurement apparatus exports metadata related to test data at 30. The metadata may include user input, parameters of the oscilloscope and the probe, the state of the technology such as standards and versions, voltage levels, current, temperature, and other operating parameters. Also, the oscilloscope imports data into the database 14 at 32. Raw waveforms and user interactions are automatically labeled at 34. The automatic labeling classifies and deletes unnecessary data that is not useful for suggesting further analysis. The automatically labeled data is also sent to the database 14. The system may adopt efficient storage management by deleting unnecessary data, compressing and serializing it in real time, and storing it on the cloud, on the device, or on one or more databases. The learned weights are portable and may include version management within the organization.

[0011] The firmware of the oscilloscope receives sampled analog data, so the labeling at the firmware level regarding the characteristics of the data is minimized. These labels may include the sample rate, time / frequency domain data, the technology format of the test object, data based on the probe used, etc. All data received by the oscilloscope is classified based on the metadata of the usage pattern and the user's suggestions to the model. The labeling system according to some embodiments of the present disclosure may have a supervision block that performs weak labeling of the data using probabilistic values.

[0012] For each technology (technical form) to be analyzed, there may be a classifier hosted on the oscilloscope. As an example, using a sequential ensemble approach, each technology classifier can be combined to reach a final classification. Provide proposals for measurement, plotting, and predictive analysis of waveform data using the waveform data classified into a probable technology.

[0013] Returning to Figure 1, this information is also used at 36 to determine whether sufficient data is available to select an appropriate model. This architecture may have a function at 38 to create a model using a plugin or a pre-defined program. Returning to 36, if sufficient data is not available, the process waits until more data becomes available, which is indicated by the return path to the oscilloscope. If an appropriate model is available, the neural network 18 executes an operation at 40 to select a model suitable for the neural network 18.

[0014] The following pseudocode shows one embodiment of feature extraction from data, data sample collection, and model selection. It also shows the verification and the determination of whether the model is ready to be used or more data is needed. Wfm[1 : n] = Acquire (Ch[1 : n]) Data_Cleaning() %% Preprocessing MetaData = Extract_MetaData(Wfm[1 : n]) %% Settings, signals, GaN, temperature from the waveform [[ID=l7]]%% Additional information such as [ feat1, feat2, … featn] = Extract_Features(Wfm[1 : n]) Wait_For_Enough_Samples_To_Start_Modelling() %% Collect data samples when the user is using an oscilloscope. This is done until enough samples are obtained to generate a model. structModel = ML_Modelling(feat1, feat2, … featn, MetaData) Decide_Number_Of_Layers_BasedOnTechnology MultiLayer_DeepNetwork() %% LMS convolutional neural network, RELU, linear convolutional pre-trained model of MATLAB® Decide_Model_To_Choose %% Automatically select a model based on the technology format and complexity. The optimal model is selected based on the training data and cross-validation data. Cross_Validation_Test() Model_Results_With_Confidence_Percentage() Confusion_MatrixPlot() RetrainModel () %% If there is a gap between the prediction result and the measurement result, use the set of that data and features to fine-tune the model parameters. Model_ReadyToUse() Provide_Additional_Info_With_GraphicalPlots() Required_Waveform_DataReady()

[0015] If a suitable model is available at 42, the model operates based on the data collected by the test measurement device, provides prediction details at 44, and generates graphical and scalar results at 46. When the model has enough data to learn and operate on its own, the AI assistant interface 50 becomes available on the interface of the test measurement device. When the user interacts with the AI assistant, the AI assistant may use the large language model LLM at 52, take actions at 56, and update the knowledge database at 54.

[0016] This system architecture has the flexibility to store the elements of the system on the test measurement device, in the cloud, or on another computing device connected to the test measurement device. These elements of the system may exist in the cloud and have the cloud designation 16.

[0017] Figure 2 shows an embodiment of the measurement settings for training an AI assistant. The first panel 60 shows a settings screen with the learning process turned off. In the following screens, the selected options have thick black lines surrounding them, as shown by surrounding the off areas in panel 60. The second panel 62 shows that the setting is on and the user has selected the automatic setting. Thereby, the system trains the AI assistant in a pre-set automatic manner. Panel 64 shows some specified embodiments for custom learning. The user has set the data distribution ratio between the training set and the cross-validation set to 80% for training. The user can set the parts of the data set to be used for training and validation. The user has also selected a set of labels for the data for the automatic labeling process described above. At the bottom of the panel where the training of the AI assistant is on, the model import option can be used to import the model from the model storage, which may occur when the model already exists and re-training or fine tuning is being performed. The user can also export the model when re-training or fine tuning is completed.

[0018] In the following figures, the situation where the AI assistant is available is displayed as an AI assistant icon 70 in Figure 3, which is represented in both the screen and the exploded view. In some embodiments, there may be various options for the user's interaction with the AI assistant. For example, the embodiment may utilize three known inputs from the mouse, double-click, single-click, or right-click. Different responses can be obtained from the AI assistant for each of these inputs. In one option, when the user selects any of the above inputs (e.g., double-click), the AI assistant may automatically execute the action indicated at the position of the AI assistant icon. In another option (e.g., single-click), the AI assistant can enter the "attendant" mode, in which case the AI assistant guides the user through the procedure for executing the action instructed by the user. Also, in another option (e.g., right-click), the AI assistant can open a chatbot to allow the user to ask questions. This would involve the above-mentioned LLM so that the AI assistant can "understand" the questions and provide answers.

[0019] As a first example of interaction with the AI assistant, the AI assistant indicates that it is necessary to change the current settings in order to acquire a waveform (acquire waveform data) and perform a measurement. This problem can be seen in the output window 76. In this description, the output takes the form of a waveform, but the waveform is not displayed. The appearance of the AI assistant icon 70 in the measurement badge "Measurement 1" 72 indicates that the AI assistant has a solution for the empty input. Similarly, the icon of the AI assistant is displayed in the Ch1 badge 74. The user may click on any of the icons in any of the three input modes described above. It should be noted that in the display of a typical test measurement device, there are many other items displayed than those shown in FIGS. 3 to 10. These items are not shown in these figures for the sake of simplifying the display, but this does not mean that these items do not exist in the following figures.

[0020] Figure 4 shows the AI assistant entering the "accompanying mode". In this mode, the AI assistant can guide the user through the process of modifying the measurement settings. The AI assistant launches a settings window 77 that shows the current settings previously selected by the user. The AI assistant indicates the settings that need to be changed. In the settings window, the AI assistant indicates that the user needs to select the setting button 79 highlighted with a dashed line. Also, the AI assistant indicates at 75 that the user needs to change the setting from Vds to Id. Figure 4 also shows in the lower bar the badges that display Ch1 of 74, reference 2 of 73, and reference 3 of 80. Note that these windows are not shown in later figures.

[0021] Figure 5 shows the change of settings in 75. The partially hidden output window 76 indicates that a waveform is currently appearing. Figure 6 shows another example of interaction with the AI assistant. In this example, the user has selected to chat with the AI assistant shown at 82.

[0022] Figure 7 shows the window 84 that enables the user to select a chat with the AI assistant. Figure 8 shows the chat window 86 as part of the display. The chat window enables the AI assistant to make a proposal 87 and provides a way for the user to take an action to implement the proposal. In the example shown in Figure 8, the AI assistant recommends to the user to add a DQ0 plot to assist with the analysis. The AI assistant also provides a button 88 for performing the addition.

[0023] Figure 9 shows the resulting DQ0 plot and the phasor diagram 90. Note in Figure 9 that the central window of the display in Figure 8 has moved to make more space available for the phasor diagram and the chat window. The display shows that the central window is still being displayed.

[0024] As described above, the AI assistant has a function of predicting the results of measurements that the test measurement device did not actually perform or calculate. Figure 10 shows this example. As can be seen in the harmonic / FRA (Frequency Response Analysis) result window 89, the test measurement device calculates the index, frequency, gain, linear gain, and phase for a plurality of indices 32 to 43. As indicated by the AI assistant icon adjacent to indices 44 to 54, the AI assistant has predicted these measured values, and the test measurement device does not actually need to perform these calculations. The AI assistant extrapolates these data from previous measured values (extrapolate: estimate unknown data beyond the range of known data from known data). This allows the user to save a significant amount of time. In reliability tests, very long events, in some cases exceeding one million events, are captured, which requires test times of tens of thousands of hours and various operating parameters are calculated. The approach using this embodiment is beneficial for predicting with fewer data sets, as shown above, and is not limited to one measurement but can be extended to multiple measurements.

[0025] Figures 3 to 10 show examples of some functions of the AI assistant. These examples are not intended to show all the functions of the AI assistant, but are merely intended as examples.

[0026] Figure 11 shows a flowchart of an embodiment of the AI assistant. Similar to the system architecture of Figure 1, the process starts with the capture of test data. In the example of Figure 11, the test data takes the form of waveform 100. Waveform 100 and settings 102 are used to perform measurements 1 to n at 104 to 106. The waveform undergoes preprocessing and cleaning 110 of the data. Results 107 (i.e., results 1 to n of measurements 1 to n) and other information constitute metadata 108. As described above, the data undergoes weak labeling (labeling process) at 112 and feature extraction 114 as part of the automatic labeling process.

[0027] The labeling system includes a supervision block for performing weak labeling with probability values. For each technology (technical format) to be analyzed, the test measurement device may host a classifier. As an example, the sequential ensemble method can combine each technology classifier to reach the final classification. By using the test data classified into "almost certain" technologies, the AI assistant can propose predictive analysis of measurement values, plots, and test data.

[0028] Returning to the flowchart, at 116, the process determines whether there are sufficient samples. If no, the process continues and waits at 118. At 116, if there are sufficient samples, the process creates a model for the first process or re-learns / fine-tunes the model at 120. Thereafter, the model is verified at 122. If the predicted value matches the actual value at 124, the AI assistant can display the results at 126 and generate other available information. If the matching result is unacceptable at 124, the model is fine-tuned at 128 until the result at 124 is acceptable.

[0029] As used in this application, the term "test data" refers to data based on tests represented by waveforms. As used in this application, the term "operation data" refers to test data after preprocessing / cleaning, automatic labeling, and feature extraction, and the model operates based on this data. Data related to tests including test settings, measurement values, results, and possibly others constitutes metadata.

[0030] The operation data and metadata are stored in the database 14. The user can select a data storage format between the database 14 and the cloud 16 at 130. The data storage provides the user interface 50 in FIG. 1. In this example, as a chatbot, it uses the LLM at 52, stores information in the knowledge database 54, and takes actions at 56. This flowchart constitutes an embodiment of a process flow as an example. Note that it is not intended to limit or imply a specific implementation form or order of these processes.

[0031] As described above, the user can set a part of the learning data collected during operation between learning and verification. The test measurement device can provide plots such as an efficiency plot, a loss plot, and a histogram of the performance of the machine learning model based on analysis. FIG. 12 shows an efficiency plot between the learning accuracy 132 and the verification accuracy 134.

[0032] In this way, the AI assistant extends the functions of the test measurement device. The advantages include shortening the test time and providing an interpretation of the predicted measurement results to designers and verification engineers. The user can easily obtain additional insights such as DUT parameters such as the product life of the DUT, such as the time to failure. This is beneficial for validation engineers to perform predictive maintenance. As described above, the system of the present application does not change the existing customer workflow, and the proposed algorithm learns on the spot when the user continues to use the oscilloscope during design and verification. According to this model, additional insights can be obtained, such as the time point when a failure occurs and the judgment of selecting the optimal filter from a large set.

[0033] Aspects of the disclosed technology can operate on a specially programmed general-purpose computer that includes specially created hardware, firmware, a digital signal processor, or a processor that operates according to programmed instructions. The terms "controller" or "processor" in this application are intended to include microprocessors, microcomputers, ASICs, and dedicated hardware controllers, among others. Aspects of the disclosed technology can be realized by computer-usable data and computer-executable instructions, such as one or more program modules, that are executed by one or more computers (including monitoring modules) and other devices. Generally, program modules include routines, programs, objects, components, data structures, etc., which, when executed by a processor within a computer or other device, perform a particular task or implement a particular abstract data type. The computer-executable instructions may be stored on a computer-readable storage medium such as a hard disk, optical disk, removable storage medium, solid-state memory, RAM, etc. As will be understood by those skilled in the art, the functions of the program modules may be combined or distributed as necessary in various embodiments. Further, such functions may be embodied in whole or in part in firmware or hardware equivalents such as integrated circuits, field-programmable gate arrays (FPGAs), etc. One or more aspects of the disclosed technology can be more effectively implemented using a particular data structure, and such data structures are considered to be within the scope of the computer-executable instructions and computer-usable data described in this application.

[0034] The disclosed embodiments may, in some cases, be implemented in hardware, firmware, software, or any combination thereof. The disclosed embodiments may be implemented as instructions carried or stored by one or more computer-readable media readable by and executable by one or more processors. Such instructions may be referred to as a computer program product. As used herein, a computer-readable media means any media accessible by a computing device. By way of example and not limitation, a computer-readable media may include computer storage media and communication media.

[0035] Computer storage media means any media that can be used to store computer-readable information. By way of example and not limitation, computer storage media may include random access memory (RAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technologies, compact disc read only memory (CD-ROM), digital versatile disc (DVD) or other optical disc storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, and any other removable or non-removable media implemented with any technology for storing volatile or nonvolatile information. Computer storage media excludes signals per se and transient forms of signal transmission.

[0036] Communication media means any media that can be used to communicate computer-readable information. By way of example and not limitation, communication media may include coaxial cable, fiber optic cable, air, or any other media suitable for communicating electrical, optical, radio frequency (RF), infrared, acoustic, or other types of signals. Examples

[0037] Examples that are useful for understanding the technology disclosed in this application are presented below. Embodiments of this technology may include one or more of the examples described below and any combination thereof.

[0038] Example 1 is a test measurement device, one or more ports for connecting to a device under test (DUT), a user interface having one or more operation units, a display, a storage, one or more processors and, the one or more processors perform a process of receiving a test signal from the DUT through the one or more ports during the test of the DUT, perform a process of generating test data using the test signal, perform a process of displaying the test data on the display unit, perform a process of displaying on the user interface an operation button indicating that an artificial intelligence (AI) assistant is available, perform a process of receiving an input from a user through the operation button to activate the AI assistant, perform a process of providing on the user interface an area where the user can interact with the AI assistant, when an input is received through the area on the user interface, perform a process of applying a machine learning model represented by the AI assistant to provide the user with additional information related to the test and one or more of the DUTs and is configured to execute a program (code) that causes the one or more processors to perform the above.

[0039] Example 2 is the test measurement apparatus of Example 1, and the program that causes the one or more processors to perform a process of providing additional information includes a process of providing a recommendation for measurement settings, a process of generating additional results for what the user has selected, a process of recommending additional devices, and a process of predicting the time until failure of the DUT or a component on the DUT, and causes the one or more processors to perform the processes.

[0040] Example 3 is the test measurement apparatus of either Example 1 or 2, and the program that causes the one or more processors to perform a process of providing the region on the user interface includes a program that causes the one or more processors to perform a process of displaying the operation button of the AI assistant at a relevant position on the display.

[0041] Example 4 is the test measurement apparatus of Example 3, and the one or more processors are further configured to execute a program for receiving an input via one of the operation buttons and performing a process in response to the input. The program that causes the one or more processors to perform the process in response to the input includes a process of automatically executing a recommended action represented by the operation button when the input includes a first input, and a process of displaying a procedure that enables the user to execute a recommended action when the input includes a second input, and a process of displaying an interactive window that enables the user to interact with the AI assistant when the input includes a third input and causes the one or more processors to perform the processes.

[0042] Example 5 is the test measurement apparatus of any one of Examples 1 to 4, and the one or more processors are further configured to execute a program that causes the one or more processors to create and train the machine learning model.

[0043] Example 6 is the test measurement device of Example 5, wherein the one or more processors are configured to create and train the machine learning model without interrupting the user's workflow while the user is using the test measurement device.

[0044] Example 7 is the test measurement device of any one of Examples 1 to 6, wherein the program for causing the one or more processors to create the machine learning model includes a process of preprocessing the data before the one or more processors collect the data composed of the test data and related metadata, a process of extracting main feature data from the data, a process of analyzing the linearity of the main feature data to select an activation function, a process of activating a part of the neural network using the activation function to construct the machine learning model, a process of training the machine learning model using a part of the main feature data, and a process of validating the machine learning model using another part of the main feature data to be performed by the one or more processors.

[0045] Example 8 is the test measurement device of Example 7, wherein the one or more processors are further configured to execute a program that causes the one or more processors to perform a process of preprocessing the data before extracting the main feature data.

[0046] Example 9 is the test measurement device of any one of Examples 1 to 9, wherein the one or more processors are further configured to execute a program that causes the one or more processors to manage the storage of the data by automatically labeling the data, classifying the data, deleting unnecessary data, compressing and serializing the data, and storing the data in real time.

[0047] Example 10 is any of the test measurement devices of Examples 1 to 9, and the one or more processors are further configured to execute a program that causes the one or more processors to perform a process of adjusting (tuning) the machine learning model during use of the test measurement device in order to keep the machine learning model up-to-date.

[0048] Example 11 is a method of using an artificial intelligence (AI) assistant in a test measurement device, a process of receiving a test signal from the DUT through a port during a test of the device under test (DUT), a process of generating test data using the test signal, a process of displaying the test data on a display, a process of displaying an operation button on the user interface indicating that an artificial intelligence (AI) assistant is available, a process of receiving an input from the user through the operation button to activate the AI assistant, a process of providing a region on the user interface through which the user can interact with the AI assistant, a process of providing the user with additional information related to the test and one or more of the DUTs using a machine learning model related to the AI assistant when an input is received through the region on the user interface and includes.

[0049] Example 12 is the method of Example 11, wherein the process of providing additional information includes a process of providing a recommendation for measurement settings, a process of generating additional results for what the user has selected, a process of recommending additional devices, and a process of predicting the lifespan of the DUT or a component on the DUT.

[0050] Example 13 is any of the methods of Example 11 or Example 12, wherein the process of providing the region on the user interface includes a process of displaying an operation button at a relevant position on the display of the test data.

[0051] Example 14 is the method of Example 13, and when an input is received via one of the above operation buttons, when the above input includes a first input, a process of executing a recommended action represented by the above operation button, and when the above input includes a second input, a process of displaying a procedure that enables the above user to execute a recommended action, and when the above input includes a third input, a process of displaying an interactive window for enabling the above user to interact with the above AI assistant further include a process of responding to the above input.

[0052] Example 15 is the method of any one of Examples 11 to 14, and further includes a process of creating the above machine learning model.

[0053] Example 16 is the method of Example 15, and the process of creating the above machine learning model includes a process of creating the above machine learning model when the machine learning model does not exist.

[0054] Example 17 is the method of Example 15, and the process of creating the above machine learning model includes a process of learning the above machine learning model in real time.

[0055] Example 18 is the method of Example 15, and the process of creating the above machine learning model is performed without interrupting the work of the above user while the above user is using the test measuring device.

[0056] Example 19 is the method of Example 15, and the process of creating the above machine learning model is a process of collecting the above test data and related metadata as data, and a process of extracting main feature data from the above data, and a process of analyzing the linearity of the above main feature data and selecting an activation function, and Activating a part of the neural network using the activation function described above and constructing the machine learning model; Using a configurable portion of the main feature data to train the machine learning model; Using another configurable portion of the main feature data to verify the machine learning model and having;

[0057] Example 20 is the method of Example 17, further comprising a process of preprocessing the data before the process of extracting the main feature data from the data.

[0058] Example 21 is the method of any one of Examples 11 to 20, further comprising a process of automatically labeling the data to classify the data, a process of removing unnecessary data, a process of compressing and serializing the data, and a process of storing the data in real time thereafter to manage the storage of the data.

[0059] Example 22 is the method of any one of Examples 1 to 21, further comprising a process of tuning the machine learning model during use of the test measurement device to keep the machine learning model up to date.

[0060] Example 23 is the method of Example 15, further comprising a process of sharing the machine learning model among a plurality of test measurement end points after the machine learning model is created.

[0061] Example 24 is the method of Example 23, wherein the process of sharing the machine learning model among the plurality of test measurement end points further includes a process of providing version management when another test measurement end point updates the machine learning model.

[0062] Example 25 is any of the methods of Examples 1 to 25, further comprising employing a subscription service that allows external users to access, tune, and optimize the machine learning model to suit their particular applications and requirements.

[0063] Additionally, the description of this application refers to specific features. It should be understood that the disclosure herein includes all possible combinations of these specific features. When a specific feature is disclosed in connection with a particular aspect or example, that feature can also be used in connection with other aspects and examples, to the extent possible.

[0064] Furthermore, when this application refers to a method having two or more defined steps or processes, these defined steps or processes may be performed in any order or simultaneously, unless the circumstances do not preclude this possibility.

[0065] All features disclosed in the specification, claims, abstract and drawings, and all steps in any disclosed method or process, may be combined in any combination, except where at least some of such features or steps are mutually exclusive combinations. Each feature disclosed in the specification, abstract, claims and drawings may be replaced by an alternative feature serving the same, equivalent or similar purpose, unless expressly stated otherwise.

[0066] Although specific embodiments of the invention have been illustrated and described for purposes of illustration, it will be appreciated that various modifications can be made therein without departing from the spirit and scope of the invention. Accordingly, the invention should not be limited except as by the appended claims. [Explanation of symbols]

[0067] 10 Test and measurement equipment 12 Device Under Test (DUT) 14 One or more databases 16. Cloud Storage 18 Neural network 20 One or more processors 22 User interface (U / I) 24 Display unit (display) 26 Storage 28 Port 50 AI assistant interface

Claims

1. A test measurement device, one or more ports for connecting to a device under test (DUT), a user interface having one or more operation units, a display, a storage, and one or more processors comprising, the one or more processors are configured to receive a test signal from the DUT through the one or more ports during testing of the DUT, generate test data using the test signal, display the test data on the display unit, display on the user interface an operation button indicating that an artificial intelligence (AI) assistant is available, receive an input from a user through the operation button to activate the AI assistant, provide on the user interface an area where the user can interact with the AI assistant, when an input is received through the area on the user interface, apply a machine learning model represented by the AI assistant to provide the user with additional information related to the test and one or more of the DUTs A test measurement device configured to execute a program (code) that causes the one or more processors to perform the above operations.

2. The program that causes the one or more processors to perform the process of providing additional information includes a process of providing recommendations for measurement settings, a process of generating additional results for what the user has selected, a process of recommending additional devices, and a process of predicting the time to failure of the DUT or a component on the DUT The test measurement device according to claim 1, comprising a program that causes the one or more processors to perform the above operations.

3. The program that causes the one or more processors to provide the area on the user interface includes a program that causes the one or more processors to display the operation button of the AI assistant at a relevant position on the display, The one or more processors are further configured to execute a program for receiving an input through one of the operation buttons and performing a process of responding to the input. The program that causes the one or more processors to perform the process of responding to the input includes when the input includes a first input, automatically executing a recommended action represented by the operation button When the above input includes a second input, a process of displaying a procedure that enables the user to execute a recommended action; When the above input includes a third input, a process of displaying an interactive window that enables the user to interact with the AI assistant; The test measurement device according to claim 1, comprising a program that causes the one or more processors to perform the above.

4. The test measurement device according to claim 1, wherein the one or more processors are configured to create and train the machine learning model without interrupting the user's workflow while the user is using the test measurement device.

5. A program that causes the one or more processors to perform the process of creating the machine learning model includes: A process of preprocessing the data before the one or more processors collect data composed of the test data and related metadata; A process of extracting main feature data from the data; A process of analyzing the linearity of the main feature data and selecting an activation function; A process of activating a part of the neural network using the activation function and constructing the machine learning model; A process of training the machine learning model using a part of the main feature data; A process of verifying the machine learning model using another part of the main feature data; The test measurement device according to claim 1, comprising a program that causes the one or more processors to perform the above.

6. The test measurement device according to claim 5, wherein the one or more processors are further configured to execute a program that causes the one or more processors to perform a process of preprocessing the data before extracting the main feature data.

7. The test measurement device according to claim 1, wherein the one or more processors are further configured to execute a program that causes the one or more processors to perform a process of managing the storage of the data by automatically labeling the data, classifying the data, deleting unnecessary data, compressing and serializing the data, and saving the data in real time.

8. The test measurement device according to claim 1, wherein the one or more processors are further configured to execute a program that causes the one or more processors to perform a process of adjusting the machine learning model during use of the test measurement device to keep the machine learning model in the latest state.

9. A method of using an artificial intelligence (AI) assistant in a test measurement device, comprising: receiving, during a test of a device under test (DUT), a test signal from the DUT through a port; generating test data using the test signal; displaying the test data on a display; displaying, on a user interface, an operation button indicating that the artificial intelligence (AI) assistant is available; receiving, from a user through the operation button, an input to activate the AI assistant; providing, on the user interface, an area through which the user can interact with the AI assistant; when an input is received through the area on the user interface, using a machine learning model associated with the AI assistant to provide the user with additional information related to one or more of the test and the DUT; A method of using an artificial intelligence (AI) assistant in a test measurement device, comprising the above steps.

10. The process of providing additional information includes: providing recommendations for measurement settings; generating additional results for what the user has selected; recommending additional devices; predicting the lifespan of the DUT or components on the DUT. A method of using an artificial intelligence (AI) assistant in a test measurement device according to claim 9, including the above steps.

11. The process of providing the area on the user interface includes displaying an operation button at a relevant position on the display of the test data. When an input is received through one of the operation buttons: if the input includes a first input, executing a recommended action represented by the operation button; if the input includes a second input, displaying a procedure that enables the user to execute the recommended action; if the input includes a third input, displaying an interactive window for the user to interact with the AI assistant. A method of using an artificial intelligence (AI) assistant in a test measurement device according to claim 9, further including responding to the input as described above.

12. collecting the test data and related metadata as data; extracting main feature data from the data; analyzing the linearity of the main feature data and selecting an activation function. ​ Activating a part of the neural network using the activation function and constructing the machine learning model; Using the modifiable part of the main feature data to train the machine learning model; Using another modifiable part of the main feature data to verify the machine learning model; The method of using an artificial intelligence (AI) assistant in the test and measurement device according to claim 9, further comprising a process of creating the machine learning model having the above.

Images