Control Program Replication for Hardware Fault Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing hardware fault detection methods in computer-based systems, particularly in autonomous machines, are inefficient and incur significant overhead, failing to effectively address random faults caused by unpredictable phenomena like neutron bombardment and electromagnetic interference.
Innovation Solution
Implementing multiple instances of control programs that execute simultaneously and compare their outputs to detect hardware faults, using methods such as round robin data distribution and majority voting to maintain processor overhead and enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hardware design modifications such as error correcting codes and redundant hardware blocks are employed to detect and compensate for faults, then system reliability is improved, but area, resource, and power overhead increase significantly
Solution Approach 1:
The patent creates virtual copies of the control program (multiple instances) rather than physical hardware copies. These software instances are executed on the same hardware platform, eliminating the need for redundant physical hardware blocks while maintaining fault detection capability through comparison of instance outputs
Solution Approach 2:
The patent replaces mechanical/hardware-based fault detection mechanisms (ECC, parity bits, redundant hardware) with a software-based approach using multiple program instances and output comparison. This substitution reduces hardware overhead while achieving similar or superior reliability
2Reliability
If multiple instances of control programs are executed simultaneously to detect hardware faults, then fault detection accuracy is improved, but processor overhead increases
Solution Approach 1:
The patent segments the control program into multiple independent instances that can be executed in parallel. Each instance processes the same sensor data independently, and their outputs are compared for fault detection. This segmentation enables efficient utilization of multi-core processors and reduces per-instance computational overhead
Solution Approach 2:
The patent merges multiple program instances onto a single hardware platform, sharing common resources such as memory, I/O interfaces, and processing units. This consolidation reduces overall resource consumption compared to having separate physical systems, while still achieving fault detection through output comparison
3Reliability
If redundant hardware blocks and repeated processing steps are used for fault detection, then fault coverage is improved, but area and resource overhead increase
Solution Approach 1:
The patent makes the control program instances universal by designing them to handle multiple functions: normal control operation, fault detection through self-comparison, and potential fault isolation. The same software instances serve both operational and diagnostic purposes, eliminating dedicated fault detection hardware
Data Source
AI summary
Systems and methods for detecting hardware faults in computer-based feedback control systems. Multiple instances of the system control program(s) are run on system processors. System sensor data are input to each instance, and the control commands output by each instance are compared. As instantiations of the same programs receive largely the same sensor data, differences between output commands may indicate the presence of one or more hardware faults.