Autonomous Vehicle Dependency Modeling for Dynamic Node Reconfiguration

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Autonomous vehicle (AV) controllers face optimization challenges due to hardware and software module changes, leading to suboptimal resource utilization and performance, as existing optimizations become invalid with upgrades or replacements, resulting in serialized operations and decreased efficiency.

Innovation Solution

A global optimizer system dynamically reconfigures AV software and hardware settings by maintaining a dependency graph, iteratively testing resource assignments, and applying machine learning optimizations to ensure optimal performance and efficiency, even with changes in hardware or software configurations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If hardware or software modules are upgraded or replaced in AV controllers, then system capabilities and performance are improved, but existing optimizations become invalid and require complete re-optimization

Engineering Contradiction:

Improvehardware adaptabilityVSAvoidoptimization complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system dynamically re-optimizes resource assignments and operational parameters whenever hardware or software changes are detected. The global optimizer continuously monitors configuration changes and automatically adjusts compute node assignments, data flow paths, and resource allocations to maintain optimal performance without manual intervention.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements a feedback mechanism where performance metrics are continuously collected from compute nodes and resource utilization data. This feedback loop enables the global optimizer to detect when optimizations have become suboptimal due to hardware/software changes and triggers automatic re-optimization cycles to restore peak performance.

Inventive Principle:

Principle #23Feedback

2Productivity

If compute nodes operate in parallel to improve processing speed, then productivity increases, but resource contention and coordination complexity increase

Engineering Contradiction:

Improveprocessing speedVSAvoidcoordination complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system segments compute nodes into specialized groups based on their functional roles (e.g., perception nodes, prediction nodes, planning nodes). Each segment handles specific autonomous driving tasks independently, reducing coordination overhead while maintaining parallel processing efficiency. The segmentation allows nodes within each group to operate with minimal inter-dependency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The global optimizer acts as an intermediary that manages resource allocation and data flow between parallel compute nodes. It maintains a centralized view of system state and coordinates resource assignments, resolving contentions before they occur and ensuring efficient data exchange between nodes without requiring complex peer-to-peer coordination protocols.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If resource allocation is optimized for specific hardware configurations, then performance is maximized, but the system cannot adapt when hardware changes occur

Engineering Contradiction:

Improveperformance efficiencyVSAvoidhardware adaptability

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The system dynamically changes operational parameters including compute node assignments, data flow configurations, and resource allocation strategies based on detected hardware or software changes. When modifications are identified, the global optimizer adjusts these parameters to re-establish optimal performance for the new configuration without requiring manual reconfiguration.

Inventive Principle:

Principle #35Parameter changes

4Productivity

If manual optimization is performed for each hardware configuration, then performance is optimized, but time and labor requirements increase significantly

Engineering Contradiction:

Improveperformance optimizationVSAvoidoptimization time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs self-optimization through an automated global optimizer that continuously monitors system state, detects hardware/software changes, and automatically re-configures resource assignments. This eliminates the need for manual optimization intervention, reducing both time and labor requirements while maintaining peak performance across configuration changes.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS20230406352A1Dependency modeling for autonomous vehicles

Publication Date: 2023.12.21 GM CRUISE HOLDINGS LLC

AI summary

A computing apparatus, comprising: a processor circuit and a memory; and instructions encoded within the memory to instruct the processor to: receive a stored dependency graph for a first hardware configuration for a plurality of compute nodes; receive a second hardware configuration comprising a modification to the first hardware configuration; iteratively model the second hardware configuration comprising adjusting one or more of a plurality of operational parameters for a model of the second hardware configuration, and selecting an optimum configuration of the plurality of operational parameters; and sending the optimum configuration to a real-world embodiment of the second hardware configuration.