There is provided a substrate inspection apparatus capable of inspecting the electrical characteristics of a packaged semiconductor device in a mounting environment. A prober includes a test box, a probe card and a package inspection card. A packaged device is attached to the package inspection card. A test board of the test box and a card board of the package inspection card reproduce the mounting environment in which a wafer-level system-level test is performed.

A data intersection is assessed of data to be used between at least two parties. The data is to be used in an artificial intelligence (AI) application. Evaluation is performed of set of instructions required for the AI application, where the evaluation creates a modified set of instructions where operands are symbolically associated with corresponding privacy levels. Using the assessed data intersection and the modified set of instructions, a mapping is created from the data to operands with associated privacy metrics. The mapping treats overlapping data from the assessed data intersection differently from data that is not overlapping to improve privacy relative to without the mapping. The AI application is executed using the data to produce at least one parameter of the AI application. The at least one parameter is output for use for a trained version of the AI application. Apparatus, methods, and computer program products are described.

A recording medium recording a program for a process, the process includes: calculating an amount of distortion in a via of a printed circuit board based on an expression using coefficient m, ={(LtE)/(DT)}m, where is the amount of distortion, L is a length of the via, is a thermal expansion coefficient of a base material, t is a temperature change of an environment, E is a Young's modulus, D is a diameter of the via, and T is a thickness of plating in the via; and calculating a lifetime of the via based on an expression, M=N/(n365), where M is the lifetime of the via, n is a frequency of the temperature change, and N is the number of cycles of the lifetime satisfying an expression N.sup.x=C/.

A three-phase diagnosis methodology capable of effectively diagnosing and classifying multiple defects in integrated circuits comprises a first phase identifying a defect that resembles traditional fault models, and second and third phases that utilize the X-fault model and machine learning to identify correct candidates.

Methods and systems are described to identify potential failures caused by metastability arising from signal propagation between asynchronous clock domains in integrated circuits with multiple operating modes, each mode allowing selected clocks to propagate. Typical integrated circuits have numerous operating modes, and hence numerous possible clock combinations, each combination causing different asynchronous clock-domain crossings, and hence different potential failures. Since verification for even one clock combination is time-consuming, explicitly enumerating and verifying all possible clock combinations is unviable. In practice very few clock combinations are verified, possibly missing failures. The present invention achieves superior performance, scalability, comprehensiveness and precision in verification despite numerous operating modes, due the following insights: (a) The number of possible clock combinations for a transmit-receive signal pair is small relative to the total number of operating modes, and (b) Cause of failure for a transmit-receive pair remain identical across many clock combinations associated with it.

In example implementations described herein, the power of time series machine learning is used to extract the statistics of Programmable Logic Controller (PLC) data and external sensor data. The accuracy of time series machine learning is improved by manufacturing context-dependent segmentation of the time series into states which is factory may be in. The invention can capture subtle trends in these time series data and be able to classify them into several outcomes from ICS security attacks to normal anomalies and machine/sensor failures.

An example method comprises receiving historical sensor data from sensors of components of wind turbines, training a set of models to predict faults for each component using the historical sensor data, each model of a set having different observation time windows and lead time windows, evaluating each model of a set using standardized metrics, comparing evaluations of each model of a set to select a model with preferred lead time and accuracy, receive current sensor data from the sensors of the components, apply the selected model(s) to the current sensor data to generate a component failure prediction, compare the component failure prediction to a threshold, and generate an alert and report based on the comparison to the threshold.

Systems and methods of testing the health of vehicular power devices are disclosed herein. A method may include producing operating points as a function of cycling current (I.sub.ds) and voltage drain to source (V.sub.ds) when a subject device is conducting current. The method may further include determining a mean of moving distribution to adapt a center of the moving distribution contrasted with a plurality of known healthy devices. The method may also include indicating an imminent fault in the subject device based upon a discontinuity among operating points above a threshold.

System and method for autonomous trouble shooting of a unit under test (UUT) having a plurality of replaceable components include: a test station that stores an artificial intelligence (AI) program and a knowledge database (KDB) including acceptable test results for each test point represented by an acceptable test vector, a test probe to test the circuit card assembly; and an operator station to send commands to the test station via the communication network to teach the AI program to capture and store the acceptable test result for each test point of the UUT by the test probe, in the KDB, wherein the AI program commands the test probe to test the UUT, stores the test results in a test result vector, compares the test result vector with the stored acceptable test vector, and displays recommendation as which replaceable component in the UUT to be repaired or replaced.

Solar array fault detection, classification, and localization using deep neural nets is provided. Embodiments use a cyber-physical system (CPS) approach to fault detection in photovoltaic (PV) arrays. Customized neural network algorithms are deployed in feedforward neural networks for fault detection and identification from monitoring devices that sense data and actuate at each individual module in a PV array. This approach improves efficiency by detecting and classifying a wide variety of faults and commonly occurring conditions (e.g., eight faults/conditions concurrently) that affect power output in utility scale PV arrays.