Conformal Asset State Prediction for Proactive IT Failure Remediation
Solution Overview
Problem
Current information processing systems lack the intelligence to predict asset failures proactively, leading to potential downtime and increased priority levels only after a failure state is reached, without anticipating the next state of devices and the impact of issues on the entire data center.
Innovation Solution
The implementation of conformal prediction techniques, specifically using a Markov chain and Mondrian conformal prediction, to identify the probability of asset state transitions and predict future states, allowing for proactive modification of priority levels and remediation measures before asset failures occur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional reactive monitoring is used to detect asset failures, then the system can identify issues after they occur, but the priority level is increased only after failure state is reached, causing downtime and preventing proactive remediation
Solution Approach 1:
The system performs preliminary actions by predicting future asset states before failures occur. Using conformal prediction and Markov chains, the system anticipates potential failure states and triggers remediation actions in advance, transforming the reactive monitoring approach into a proactive one that prevents downtime rather than responding to it after occurrence.
Solution Approach 2:
The system applies beforehand cushioning by preparing and executing remediation measures before asset failures manifest. By predicting future states with confidence intervals and prioritizing support indicators based on predicted outcomes, the system cushions against potential failures by having remediation ready or already applied, thereby preventing the harmful effect of complete asset failure and associated downtime.
2Reliability
If conformal prediction techniques are implemented to predict asset state transitions, then proactive remediation is enabled, but the system complexity increases due to Markov chains and probability calculations
Solution Approach 1:
The system applies self-service by automatically performing probability calculations, state predictions, and priority adjustments without requiring external analytical tools. The conformal prediction framework and Markov chain models are integrated directly into the support platform, enabling the system to serve its own prediction and remediation needs internally, thereby managing complexity through self-containment rather than external dependencies.
Solution Approach 2:
The system achieves universality by designing the conformal prediction framework to handle multiple asset types and failure modes through a single unified approach. The Markov chain models and probability calculations are applied generically across different assets, allowing the same complex machinery to serve multiple functions and asset categories, thereby justifying the complexity through broad applicability and avoiding the need for separate specialized systems.
3Measurement precision
If the system monitors all asset state transitions with high precision, then accurate predictions are achieved, but the computational resources and data processing requirements increase significantly
Solution Approach 1:
The system applies local quality by focusing computational precision only where needed - specifically on predicting future states with high confidence rather than uniformly monitoring all current states. Conformal prediction provides locally adaptive confidence intervals that concentrate computational resources on uncertain transitions while accepting lower precision on well-understood state changes, thereby reducing overall computational energy consumption while maintaining prediction accuracy where it matters most.
Data Source
AI summary
An apparatus comprises a processing device configured to identify a given one of one or more assets in an information technology infrastructure associated with a support indicator associated with a priority level, and to obtain information characterizing state transitions of the one or more assets. The processing device is also configured to determine, based at least in part on a current state of the given asset identified utilizing the obtained information, a probability of the given asset transitioning to each of a plurality of states and to select, based at least in part on the determined probabilities, one of the plurality of states as a predicted future state of the given asset utilizing conformal prediction. The processing device is further configured to modify the priority level of the support indicator associated with the given asset based at least in part on the predicted future state of the given asset.