Edge analytics for predicting software failures
Inventors
Gauf, Bernard • TOPALIS, Apostolos • HARRIMAN, Keith
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12360884-B2
Publication Date
2025-07-15
Expiration Date
Abstract
An embodiment of the present invention is directed to a novel approach of predicting software failures while executing in an operational or production environment. The innovative method and system provides analytic capabilities that monitor a system and input conditions and further provides a prediction mechanism to anticipate a software failure and present an improved course of action.
Core Innovation
The invention provides a computer implemented system that implements edge analytics to predict software failures. The system monitors a computer server during operational execution in an operational environment and executes a prediction mechanism to anticipate a software failure. A decision boundary is generated responsive to the prediction mechanism to provide a baseline to avoid the one or more software failures.
The prediction mechanism is structured as a first phase and a second phase. In the first phase, one or more failure patterns are captured from a corresponding simulated environment, and the simulated environment generates data for a relationship model that identifies one or more relationships between a first set of input conditions and test results through condition sampling. The first phase also collects operational edge analytics data across multiple deployed instances during the operational execution in the operational environment.
In the second phase, the one or more failure patterns are applied in the operational environment to identify one or more input conditions and analysis results of one or more software failures. The second phase produces a dataset and makes an assessment independent of the one or more input conditions enabling evaluations in an untested condition space. Correlations between a second set of input conditions and the one or more software failures are identified through statistical analysis, wherein the relationship model from the first phase is applied to the operational environment in the second phase.
Responsive to the anticipated software failure, the system generates at least one recommended course of action.
Claims Coverage
The partial content includes two independent claims. Across these claims, the inventive framework contains a two-phase prediction mechanism using failure patterns from a simulated environment and applying a relationship model to operational edge analytics, then producing a decision boundary baseline and generating a recommended course of action. The claims cover both a system and a corresponding method.
Edge analytics system predicting software failures using simulated failure patterns and an operational relationship model
A computer implemented system that implements edge analytics to predict software failures, comprising a memory component and a computer server executing in an operational environment and coupled to the memory component, where the server monitors during operational execution; executes a prediction mechanism operating a first phase capturing one or more failure patterns sourced from a corresponding simulated environment generating data for a relationship model through condition sampling and collecting operational edge analytics data across multiple deployed instances; and a second phase applying the one or more failure patterns in the operational environment to identify input conditions and analysis results, produce a dataset, make an assessment independent of the one or more input conditions enabling evaluations in an untested condition space, and identify correlations through statistical analysis wherein the relationship model from the first phase is applied to the operational environment; generates a decision boundary baseline to avoid the one or more software failures; and generates at least one recommended course of action.
Edge analytics method predicting software failures using simulated failure patterns and an operational relationship model
A computer implemented method implementing edge analytics to predict software failures comprising monitoring a computer server during operational execution; executing a prediction mechanism operating a first phase capturing one or more failure patterns sourced from a corresponding simulated environment generating data for a relationship model through condition sampling and collecting operational edge analytics data across multiple deployed instances during the operational execution in an operational environment; and a second phase applying the one or more failure patterns in the operational environment to identify input conditions and analysis results, produce a dataset, make an assessment independent of the one or more input conditions enabling evaluations in an untested condition space, and identify correlations through statistical analysis wherein the relationship model from the first phase is applied to the operational environment; generating a decision boundary baseline to avoid the one or more software failures; and generating at least one recommended course of action.
Both independent claims cover an edge-analytics prediction framework with a two-phase mechanism: failure patterns sourced from a simulated environment are used to generate a relationship model and then applied in an operational environment to identify input-condition correlations in an untested condition space. Each claim further requires generating a decision boundary baseline to avoid anticipated failures and generating at least one recommended course of action.
Stated Advantages
Provides a baseline to avoid one or more software failures in the operational environment.
Generates at least one recommended course of action responsive to the anticipated software failure.
Documented Applications
Operational execution in an operational environment for a computer server coupled to a memory component, using monitoring and operational edge analytics across multiple deployed instances.
Interested in licensing this patent?