Dual-Controller Vehicle Control for Shared Sensing Computation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving systems with 1:1 backup redundancy controllers waste computing resources and costs since the backup controller remains idle when the master controller is functioning properly.
Innovation Solution
A control method and apparatus where two controllers cooperate by sharing computing resources, with each processing different services and exchanging sensing results, allowing them to utilize each other's computing capabilities and reduce the need for high computing performance in both controllers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 1:1 backup redundancy is implemented with two autonomous driving controllers, then system reliability is improved, but computing resource utilization deteriorates due to idle backup controller
Solution Approach 1:
The backup controller is designed to perform dual functions: serving as a standby unit for reliability while simultaneously executing sensing computations when idle. This multi-functionality allows the backup controller to contribute to system operations rather than remaining completely idle, improving resource utilization while maintaining the redundancy architecture
Solution Approach 2:
The idle backup controller serves itself by performing sensing computations that do not require its primary control functions. This self-service approach allows the backup controller to utilize its computing resources productively without interfering with its backup role, reducing wasted computing power
2Reliability
If both master and backup controllers are equipped with high computing performance, then fail operational capability is improved, but system cost increases
Solution Approach 1:
The system implements differentiated computing performance allocation: the master controller maintains high computing performance for complete autonomous driving services, while the backup controller uses lower computing performance since it only needs to handle sensing computations and can take over master functions when needed. This local quality approach reduces overall system cost while maintaining fail operational capability
Solution Approach 2:
The backup controller performs only partial sensing computations rather than complete autonomous driving processing. This partial action is sufficient for its role of providing backup sensing data and taking over in failure scenarios, reducing the computing performance requirements and associated costs compared to equipping it with full autonomous driving capability
3Reliability
If the backup controller remains idle during normal operation, then master controller performance is maintained, but computing resource waste increases
Solution Approach 1:
The backup controller continuously performs sensing computations rather than remaining idle during normal operation. This continuous useful action ensures that the backup controller's computing resources are productively utilized throughout system operation, improving overall computing resource productivity while maintaining master controller performance through the same redundancy architecture
Data Source
AI summary
Embodiments of this application provide a control method and apparatus, and a means of transportation. The method includes: A first controller obtains a first sensing result based on data collected by a sensor in a first sensor group; a second controller obtains a second sensing result based on data collected by a sensor in a second sensor group; the first controller receives the second sensing result sent by the second controller; and the first controller sends a first control instruction to an executor based on the first sensing result and the second sensing result. Embodiments of this application may be applied to an intelligent vehicle or an electric vehicle, to help improve utilization of computing resources and reduce costs of a controller.