Machine Learning System for Storage Malfunction Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional technologies face challenges in quickly detecting malfunctions in storage systems, requiring extensive manual analysis of large amounts of unstructured data from logs and statistics, which is time-consuming and inefficient, leading to delayed identification of faulty components and increased need for customer support engineers.
Innovation Solution
A machine learning system that creates bitmap images from storage system data, trains convolutional neural networks to detect malfunctions, and provides a visual representation of component health, enabling rapid identification of faulty components and their severity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis of logs and statistics is used to detect malfunctions, then detection accuracy can be maintained through expert judgment, but detection time increases significantly and productivity decreases
Solution Approach 1:
The patent replaces manual mechanical analysis of logs with an automated machine learning system that processes unstructured data. The ML model learns patterns from historical logs and statistics, automatically detecting malfunctions without human intervention, thus maintaining accuracy while dramatically improving detection speed.
Solution Approach 2:
The patent introduces an intermediate machine learning model that acts as a mediator between raw log data and final malfunction detection. The model processes and interprets unstructured data, transforming it into actionable insights, thereby eliminating the need for manual analysis while preserving detection accuracy.
2Reliability
If extensive manual analysis of unstructured data is performed, then comprehensive malfunction detection is achieved, but time consumption increases and operational efficiency decreases
Solution Approach 1:
The patent performs preliminary training of the machine learning model using historical log data and statistics before actual malfunction detection. This preliminary action enables the model to learn patterns and relationships in the data, allowing it to quickly and comprehensively detect malfunctions without requiring extensive manual analysis at runtime.
Solution Approach 2:
The patent substitutes manual analysis of unstructured data with an automated ML-based processing system. The model efficiently handles large volumes of unstructured log data, maintaining comprehensive detection capability while reducing analysis time from hours or days to minutes or seconds.
3Difficulty of detecting and measuring
If more customer support engineers are deployed to analyze system data, then detection capability improves, but operational cost increases
Solution Approach 1:
The patent implements a self-service system where the machine learning model autonomously analyzes system logs and detects malfunctions without requiring human support engineers. The system serves itself by automatically processing data, generating alerts, and identifying issues, thereby eliminating the need for extensive human intervention while maintaining or improving detection capability.
Solution Approach 2:
The patent replaces human support engineers with an automated machine learning system. The ML model performs the analytical functions previously requiring human expertise, scaling detection capability without proportionally increasing the number of personnel needed, thus reducing operational costs while maintaining or enhancing detection effectiveness.
4Measurement precision
If conventional log analysis methods are used, then detailed inspection of system state is possible, but the complexity of processing unstructured data increases significantly
Solution Approach 1:
The patent replaces complex manual processing of unstructured data with a machine learning system specialized in handling such data. The ML model automatically parses, understands, and extracts meaningful information from unstructured logs, maintaining detailed inspection capability while simplifying the processing complexity through automated pattern recognition.
Solution Approach 2:
The patent transforms unstructured log data into structured representations that the machine learning model can efficiently process. By changing the parameter representation of the data (from raw text to structured features), the system maintains detailed inspection capability while reducing processing complexity through standardized data transformation.
Data Source
AI summary
Techniques for detecting state changes in a system may include receiving a first neural network that is trained to detect when the system transitions into a first resulting state, wherein the system transitions into at least a first intermediate state prior to transitioning into the final resulting state; training the first neural network using a first plurality of inputs denoting the system in the first intermediate state; obtaining a plurality of sets of internal state information of the first neural network, each set of the plurality of sets denoting an internal state of the first neural network at a different point in time after the first neural network has processed at least a portion of the first plurality of inputs; and training a second neural network, using the plurality of sets of internal state information, to detect the first intermediate state.