Abstract: Systems, methods, and other embodiments associated with computer distress-call detection and authentication are described. In one embodiment, a method includes detecting a human voice in audio content of a radio signal. Speech is recognized in the human voice to transform the human voice into text and vocal metrics. Feature scores are generated that represent features of the recognized speech based at least in part on the vocal metrics. The human voice is then classified as either a hoax distress call or an authentic distress call based on the feature scores. An alert is then presented indicating that the human voice is one of the hoax distress call or the authentic distress call.

Abstract: A model-based approach to determining an optimal configuration for a data center may use an environmental chamber to characterize the performance of various data center configurations at different combinations of temperature and altitude. Telemetry data may be recorded from different configurations as they execute a stress workload at each temperature/altitude combination, and the telemetry data may be used to train a corresponding library of models. When a new data center is being configured, the temperature/altitude of the new data center may be used to select a pre-trained model from a similar temperature/altitude. Performance of the current configuration can be compared to the performance of the model, and if the model performs better, a new configuration based on the model may be used as an optimal configuration for the data center.

Abstract: Detecting a counterfeit status of a target device by: selecting a set of frequencies that best reflect load dynamics or other information content of a reference device while undergoing a power test sequence; obtaining target electromagnetic interference (EMI) signals emitted by the target device while undergoing the same power test sequence; creating a sequence of target kiviat plots from the amplitude of the target EMI signals at each of the set of frequencies at observations over the power test sequence to form a target kiviat tube EMI fingerprint; comparing the target kiviat tube EMI fingerprint to a reference kiviat tube EMI fingerprint for the reference device undergoing the power test sequence to determine whether the target device and the reference device are of the same type; and generating a signal to indicate a counterfeit status based at least in part on the results of the comparison.

Abstract: During operation, the system uses N sensors to sample an electromagnetic interference (EMI) signal emitted by a target asset while the target asset is running a periodic workload, wherein each of the N sensors has a sensor sampling frequency f, and wherein the N sensors perform sampling operations in a round-robin ordering with phase offsets between successive samples. During the sampling operations, the system performs phase adjustments among the N sensors to maximize phase offsets between successive sensors in the round-robin ordering. Next, the system combines samples obtained through the N sensors to produce a target EMI signal having an EMI signal sampling frequency F=f×N. The system then generates a target EMI fingerprint from the target EMI signal. Finally, the system compares the target EMI fingerprint against a reference EMI fingerprint for the target asset to determine whether the target asset contains any unwanted electronic components.

Abstract: Techniques for using machine learning model validated sensor data to generate recommendations for remediating issues in a monitored system are disclosed. A machine learning model is trained to identify correlations among sensors for a monitored system. Upon receiving current sensor data, the machine learning model identifies a subset of the current sensor data that cannot be validated. The system generates estimated values for the sensor data that cannot be validated based on the learned correlations among the sensor values. The system generates the recommendations for remediating the issues in the monitored system based on validated sensor values and the estimated sensor values.

Abstract: Systems, methods, and other embodiments associated with detecting feedback control instability in computer thermal controls are described herein. In one embodiment, a method includes for a set of dwell time intervals, wherein the dwell time intervals are associated with a range of periods of time from an initial period to a base period, executing a workload that varies from minimum to maximum over the period on a computer during the dwell time interval; recording telemetry data from the computer during execution of the workload; incrementing the period towards a base period; determining that either the base period is reached or a thermal inertia threshold is reached; and analyzing the recorded telemetry data over the set of dwell time intervals to either (i) detect presence of a feedback control instability in thermal control for the computer; or (ii) confirm feedback control stability in thermal control for the computer.

Abstract: Systems, methods, and other embodiments associated with unified control of cooling in computers are described. In one embodiment, a method locks operation of first and second cooling mechanisms configured to cool one or more components in the computer. In response to a first condition, the method unlocks the operation of the first cooling mechanism to allow the first cooling mechanism to make cooling adjustments while the operation of the second cooling mechanism is locked. In response to a second condition, the method unlocks the operation of the second cooling mechanism to allow the second cooling mechanism to make cooling adjustments while the operation of the first cooling mechanism is locked. In the method, the first cooling mechanism and the second cooling mechanism are prevented from making the cooling adjustments simultaneously.

Abstract: Systems, methods, and other embodiments associated with autonomous discrimination of operation vibration signals are described herein. In one embodiment, a method includes partitioning a frequency spectrum of output into a plurality of discrete bins, wherein the output is collected from vibration sensors monitoring a reference device; generating a representative time series signal for each bin while the device is operated in a deterministic stress load; generating a PSD for each bin by converting each signal from the time domain to the frequency domain; determining a maximum power spectral density value and a peak frequency value for each bin; selecting a subset of the bins that have maximum PSD values exceeding a threshold; assigning the representative time series signals from the selected subset of bins as operation vibration signals indicative of operational load on the reference device; and configuring a machine learning model based on at least the operation vibration signals.

Abstract: Embodiments for passive spychip detection through polarizability and advanced pattern recognition are described. For example a method includes inducing a magnetic field in a passive component of a target system while the target system is emitting EMI with changes in amplitude repeating at a time interval; generating a time series of measurements of a combined magnetic field strength of the induced magnetic field and the EMI; executing a frequency-domain to time-domain transformation on the time series of measurements to create time series signals of combined magnetic field strength over time at a specific frequency range; monitoring the time series signals with an ML model trained to predict correct signal values to determine whether predicted and measured values of the time series agree; and indicating that the target device may contain a passive spychip where anomalies are detected, and is free of passive spychips where no anomalies are detected.

Abstract: Systems, methods, and other embodiments associated with passive inferencing of signal following in multivariate anomaly detection are described. In one embodiment, a method for inferencing signal following in a machine learning (ML) model includes calculating an average standard deviation of measured values of time series signals in a set of time series signals; training the ML model to predict values of the signals; predicting values of each of the signals with the trained ML model; generating a time series set of residuals between the predicted values and the measured values; calculating an average standard deviation of the sets of residuals; determining that signal following is present in the trained ML model where a ratio of the average standard deviation of measured values to the average standard deviation of the sets of residuals exceeds a threshold; and presenting an alert indicating the presence of signal following in the trained ML model.

Abstract: Techniques for identifying a root cause of an operational result of a deterministic machine learning model are disclosed. A system applies a deterministic machine learning model to a set of data to generate an operational result, such as a prediction of a “fault” or “no-fault” in the system. The set of data includes signals from multiple different data sources, such as sensors. The system applies an abductive model, generated based on the deterministic machine learning model, to the operational result. The abductive model identifies a particular set of data sources that is associated with the root cause of the operational result. The system generates a human-understandable explanation for the operational result based on the identified root cause.

Abstract: A model-based approach to determining an optimal configuration for a data center may use an environmental chamber to characterize the performance of various data center configurations at different combinations of temperature and altitude. Telemetry data may be recorded from different configurations as they execute a stress workload at each temperature/altitude combination, and the telemetry data may be used to train a corresponding library of models. When a new data center is being configured, the temperature/altitude of the new data center may be used to select a pre-trained model from a similar temperature/altitude. Performance of the current configuration can be compared to the performance of the model, and if the model performs better, a new configuration based on the model may be used as an optimal configuration for the data center.

Abstract: The disclosed embodiments provide a system that estimates a remaining useful life (RUL) for a fan. During operation, the system receives telemetry data associated with the fan during operation of the critical asset, wherein the telemetry data includes a fan-speed signal. Next, the system uses the telemetry data to construct a historical fan-speed profile, which indicates a cumulative time that the fan has operated in specific ranges of fan speeds. The system then computes an RUL for the fan based on the historical fan-speed profile and empirical time-to-failure (TTF) data, which indicates a TTF for the same type of fan as a function of fan speed. Finally, when the RUL falls below a threshold, the system generates a notification indicating that the fan needs to be replaced.

Abstract: The disclosed embodiments relate to a system that optimizes a prognostic-surveillance system to achieve a user-selectable functional objective. During operation, the system allows a user to select a functional objective to be optimized from a set of functional objectives for the prognostic-surveillance system. Next, the system optimizes the selected functional objective by performing Monte Carlo simulations, which vary operational parameters for the prognostic-surveillance system while the prognostic-surveillance system operates on synthesized signals, to determine optimal values for the operational parameters that optimize the selected functional objective.

Abstract: The disclosed embodiments relate to a system that automatically selects a prognostic-surveillance technique to analyze a set of time-series signals. During operation, the system receives the set of time-series signals obtained from sensors in a monitored system. Next, the system determines whether the set of time-series signals is univariate or multivariate. When the set of time-series signals is multivariate, the system determines if there exist cross-correlations among signals in the set of time-series signals. If so, the system performs subsequent prognostic-surveillance operations by analyzing the cross-correlations. Otherwise, if the set of time-series signals is univariate, the system performs subsequent prognostic-surveillance operations by analyzing serial correlations for the univariate time-series signal.

Abstract: The disclosed embodiments relate to a system that automatically adapts a prognostic-surveillance system to account for aging phenomena in a monitored system. During operation, the prognostic-surveillance system is operated in a surveillance mode, wherein a trained inferential model is used to analyze time-series signals from the monitored system to detect incipient anomalies. During the surveillance mode, the system periodically calculates a reward/cost metric associated with updating the trained inferential model. When the reward/cost metric exceeds a threshold, the system swaps the trained inferential model with an updated inferential model, which is trained to account for aging phenomena in the monitored system.

Abstract: Techniques for geometric aging data reduction for machine learning applications are disclosed. In some embodiments, an artificial-intelligence powered system receives a first time-series dataset that tracks at least one metric value over time. The system then generates a second time-series dataset that includes a reduced version of a first portion of the time-series dataset and a non-reduced version of a second portion of the time-series dataset. The second portion of the time-series dataset may include metric values that are more recent than the first portion of the time-series dataset. The system further trains a machine learning model using the second time-series dataset that includes the reduced version of the first portion of the time-series dataset and the non-reduced version of the second portion of the time-series dataset. The trained model may be applied to reduced and/or non-reduced data to detect multivariate anomalies and/or provide other analytic insights.

Abstract: The disclosed embodiments relate to a system that produces anomaly-free training data to facilitate ML-based prognostic surveillance operations. During operation, the system receives a dataset comprising time-series signals obtained from a monitored system during normal, but not necessarily fault-free operation of the monitored system. Next, the system divides the dataset into subsets. The system then identifies subsets that contain anomalies by training one or more inferential models using combinations of the subsets, and using the one or more trained inferential models to detect anomalies in other target subsets of the dataset. Finally, the system removes any identified subsets from the dataset to produce anomaly-free training data.

Abstract: The disclosed system produces synthetic signals for testing machine-learning systems. During operation, the system generates a set of N composite sinusoidal signals, wherein each of the N composite sinusoidal signals is a combination of multiple constituent sinusoidal signals with different periodicities. Next, the system adds time-varying random noise values to each of the N composite sinusoidal signals, wherein a standard deviation of the time-varying random noise values varies over successive time periods. The system also multiplies each of the N composite sinusoidal signals by time-varying amplitude values, wherein the time-varying amplitude values vary over successive time periods. Finally, the system adds time-varying mean values to each of the N composite sinusoidal signals, wherein the time-varying mean values vary over successive time periods.

Abstract: Techniques for generating imputation-based, uniformly sampled parallel streams of time-series data are disclosed. A system divides into two subsets a dataset made up of multiple data streams. The data streams include interpolated data. The system trains one data correlation model using one subset of the data and applies the trained model to the other subset. The system replaces the interpolated values in the other subset with estimated values generated by the model. The system trains another data correlation model using the revised subset. The system applies the new model to the initial subset to generate estimated values for the initial subset. The system replaces the interpolated values in the initial subset with the estimated values. The system repeats the process of training data correlation models and revising previously-interpolated data points in the subsets of data until a predetermined iteration threshold is met.