Vehicle Memory Pooling for Remote AI Processing Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in memory bandwidth and component reliability due to constraints in I/O technology, power, and packaging, which can limit their performance and safety, especially when driver impairment or other factors affect vehicle operation.
Innovation Solution
A system that allows for the sharing of memory resources among physically remote entities, where a first vehicle can determine availability and grant access to processing resources or memory capacity to a second vehicle, either directly or through a third vehicle, to support operations, ensuring efficient use and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous vehicles use artificial intelligence and machine learning to improve decision-making capability, then the intelligence and safety of the vehicle are improved, but the memory bandwidth requirement increases significantly, which conflicts with I/O technology constraints and power limitations
Solution Approach 1:
The patent merges memory resources from multiple vehicles into a shared pool, allowing vehicles to access memory capacity beyond their local constraints. This combining of resources enables AI/ML operations to proceed with adequate memory bandwidth without requiring each individual vehicle to have excessive local memory, thus addressing the energy and I/O constraints while maintaining safety requirements.
Solution Approach 2:
The shared memory resource serves multiple functions: it acts as local memory for the host vehicle, provides supplemental memory for remote vehicles, and enables memory pooling across the vehicle network. This multi-functionality allows the same memory resource to be dynamically allocated to different vehicles based on their instantaneous needs, optimizing the use of memory bandwidth and reducing overall system power consumption.
2Productivity
If autonomous vehicles increase system component count to meet performance requirements, then the processing capability is improved, but the reliability decreases due to increased chance of system failure
Solution Approach 1:
The patent combines memory resources from multiple vehicles into a unified shared pool, allowing the system to achieve high processing capability through resource aggregation rather than through increasing individual vehicle component counts. This approach maintains reliability by reducing the number of redundant components needed while still meeting performance requirements through shared access to pooled memory resources.
3Reliability
If autonomous vehicles reduce system component count to decrease failure chance, then the reliability is improved, but the processing capability and memory bandwidth become insufficient
Solution Approach 1:
The patent introduces a shared memory resource pool as an intermediary between vehicles with limited local memory and AI/ML workloads requiring extensive memory bandwidth. This intermediary enables vehicles to access additional memory capacity remotely through the network, maintaining high processing capability without requiring each vehicle to have excessive local components, thus preserving reliability while enabling sufficient processing power.
4Productivity
If vehicles share memory resources over wireless network, then the memory utilization efficiency is improved, but the data transmission speed and bandwidth are limited by wireless communication constraints
Solution Approach 1:
The patent implements partial sharing of memory resources, where vehicles can access the shared pool when needed rather than requiring constant high-speed connectivity. The system allows memory access operations to proceed with appropriate protocols that account for wireless constraints, enabling sufficient memory utilization efficiency without requiring excessive data transmission speeds that would be impractical for wireless networks.
Data Source
AI summary
An example apparatus includes a first vehicle configured to determine an availability of processing resources or memory capacity, or both, at the first vehicle based at least in part on a current operating mode of the first vehicle, receive a request from a second vehicle to use at least a portion of the processing resources or the memory capacity, or both, to perform a processing operation at a second vehicle, wherein the request from the second vehicle is associated with insufficient processing capability or memory capacity, or both, at the second vehicle, and perform at least a portion of the processing operation or allow access to the available memory capacity, or both, at the first vehicle in response to the request and based at least in part on determining the availability of the processing resources or the memory capacity, or both.