Soft Error Analysis for Autonomous System Safety
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous systems, such as vehicles and machines, face safety risks due to soft errors in computing systems, which can lead to malfunctions and catastrophic consequences despite superior engineering, as they are prone to transient faults from factors like electromagnetic interference and particle radiation.
Innovation Solution
A combined soft error analysis and functional safety analysis system is implemented to determine aggregate functional safety metrics for computing systems, allowing for the identification of underperforming components and the addition of error correction codes or particle shielding to enhance safety, using predictive modeling and simulation to estimate failure rates and derive derated soft error data for improved safety metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If sophisticated computing systems are used to control autonomous vehicles, then the vehicle's navigation and obstacle detection capabilities are improved, but the risk of soft errors causing malfunctions increases
Solution Approach 1:
The system performs preliminary functional safety analysis during the design phase to identify potential soft error risks before deployment. This includes modeling soft error rates, identifying vulnerable components, and determining necessary safety measures in advance, rather than waiting for actual failures to occur.
Solution Approach 2:
The patent implements protective measures against soft errors before they can cause catastrophic failures. This includes adding error detection and correction codes, implementing redundant computing paths, and designing fail-safe mechanisms that activate when soft errors are detected, thereby cushioning the system against the full impact of potential failures.
2Measurement precision
If comprehensive functional safety analysis is performed to identify all soft error risks, then the accuracy of safety assessment is improved, but the time and complexity of analysis increases
Solution Approach 1:
The functional safety analysis is divided into distinct segments: soft error rate modeling, vulnerability factor identification, derating calculations, and safety metric determination. Each segment can be performed independently and in parallel, reducing overall analysis time while maintaining comprehensive coverage of all risk factors.
Solution Approach 2:
The system transforms raw soft error rate data into derated error rates by applying vulnerability factors as mathematical parameters. This parameter transformation approach standardizes the analysis process, enabling faster computation of safety metrics while preserving the precision needed for accurate risk assessment.
3Reliability
If derating of soft error rates is performed using multiple vulnerability factors to improve safety metric accuracy, then the reliability assessment is improved, but the computational complexity increases
Solution Approach 1:
The functional safety analysis system is designed to handle multiple types of vulnerability factors (architecture, timing, program) through a unified derating framework. This multi-functional approach allows the same system to process diverse vulnerability data using consistent mathematical relationships, reducing the need for separate analysis tools for each vulnerability type.
Solution Approach 2:
The system uses feedback loops where initial safety assessments inform subsequent vulnerability factor selections and derating adjustments. This iterative feedback process refines the reliability assessment accuracy while keeping computational complexity manageable by focusing calculations on the most significant vulnerability factors identified in previous assessment cycles.
Data Source
AI summary
Soft error data describing soft errors predicted to affect at least a particular hardware component of a computing system are used to determine functional safety metric values. The computing system is to control at least a portion of physical functions of a machine using the particular hardware component. Respective soft error rates are determined for each of a set of classifications based on the soft errors described in the soft error data. Derating of the soft error rates are performed based on a set of one or more vulnerability factors to generate derated error rate values for each of the set of classifications. The functional safety metric value is determined from the derated error rate values to perform a functional safety analysis of the computing system.