Vision Accelerator Fault Detection Using Golden Frame Signatures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices in motor vehicles face challenges in detecting faults, particularly transient and permanent faults in imaging and vision hardware accelerators, which can compromise safety and lead to unreasonable risks for drivers and passengers.
Innovation Solution
A fault detection system that includes an electronic device with a communication module, controller, hardware accelerator module, and signature generator, which obtains golden and use-case input frames, generates and compares image signatures to identify faults, ensuring compliance with safety standards like ASIL and SIL levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundancy in hardware architecture is used to detect transient faults, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent uses golden frames (pre-stored reference images) as copies of expected correct output frames. By comparing generated frames against these golden frame copies, the system detects faults without requiring redundant hardware processing paths, thus maintaining reliability while reducing complexity
Solution Approach 2:
The patent introduces an intermediary comparison mechanism that uses frame data and signature verification as a mediator between the hardware accelerator and the fault detection system. This intermediary layer enables fault detection through software-based comparison rather than direct hardware redundancy
2Reliability
If self-test mechanisms are implemented to detect permanent faults, then reliability is improved, but productivity decreases due to periodic testing interruptions
Solution Approach 1:
The patent implements periodic self-tests by inserting golden frames at regular intervals within the frame sequence. This periodic testing approach ensures reliable fault detection while maintaining continuous processing of use-case frames, minimizing productivity impact by testing only a subset of frames
Solution Approach 2:
The system performs partial testing by applying self-test mechanisms only to specific golden frames within the frame sequence, rather than testing every frame. This partial action approach maintains reliability for critical fault detection while preserving overall processing throughput
3Measurement precision
If multiple golden frames are used for comprehensive fault detection, then measurement precision is improved, but loss of time increases due to extended testing duration
Solution Approach 1:
The patent segments the fault detection process by using multiple independent golden frames, each testing specific aspects of hardware functionality. This segmentation allows comprehensive coverage of potential fault modes while enabling parallel or sequential execution that manages overall test time
Data Source
AI summary
An electronic device may be configured to detect a fault in imaging and vision hardware accelerators. The electronic device may include a controller configured to select a first golden input frame of multiple golden input frames to perform a first self-test, and retrieve a first reference image signature corresponding to the first golden input frame. The electronic device may include a hardware accelerator module configured to obtain the first golden input frame, and generate a first output frame based on the first golden input frame. The electronic device may include a signature generator configured to generate a first generated image signature based on the first output frame. The electronic device may include a signature comparison module configured to compare the first generated image signature to the first reference image signature in order to determine whether the hardware accelerator module includes a fault at a first time.