Autonomous Vehicle Input Prediction for Reliable Driving Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face errors in input data that can affect the functionality of machine learning models, leading to potential safety and operational issues.
Innovation Solution
Implementing redundant processing units, power fabrics, and data fabrics to ensure continuous operation by generating predicted input data when errors are detected, using machine learning models to handle input data errors and maintain functionality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning models are used to generate driving decisions in autonomous vehicles, then the automation capability is improved, but errors in input data may occur that affect model functionality and safety
Solution Approach 1:
The system performs preliminary actions by generating predicted input data before actual input data becomes unavailable or erroneous. Prediction models continuously generate anticipated input values that can immediately replace erroneous data, preventing interruptions in the machine learning pipeline and maintaining reliable operation during data flow disruptions.
Solution Approach 2:
Prediction models act as intermediary components between data sources and machine learning models. When input data errors occur, the prediction models provide intermediate predicted values that bridge the gap, allowing the system to maintain functionality without direct access to the original erroneous or unavailable data.
2Reliability
If redundant processing units and data fabrics are implemented to ensure continuous operation, then system reliability is improved, but device complexity increases
Solution Approach 1:
The system segments redundancy functionality into separate prediction models for different input data types rather than implementing blanket redundancy across all processing units. This segmentation allows targeted redundancy only where prediction is beneficial, reducing overall system complexity while maintaining continuous operation capability for critical data streams.
Solution Approach 2:
The system dynamically adjusts the level of redundancy and prediction based on operational parameters such as data criticality, prediction accuracy, and current system state. By changing redundancy parameters adaptively rather than maintaining fixed high redundancy, the system achieves reliable continuous operation with optimized complexity levels.
Data Source
AI summary
Processing predictive input data in an autonomous vehicle, including: receiving input data for a model; determining whether the input data for the model comprises an indication that the input data was generated based on some amount of predicted data; generating, by the model and based on the input data, output data by modifying, in response to the input data comprising the indication, one or more thresholds or one or more confidence scores of the model used in generating output data; and causing an autonomous vehicle to perform one or more driving decisions based on the output data of the model.