Dynamic Software Distribution in Vehicle Control Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of managing software redundancy in autonomous vehicles poses challenges in ensuring safety and efficiency, particularly in dynamic reconfiguration during failures, which can lead to uncontrolled driving situations and increased energy consumption.
Innovation Solution
A method for dynamic, context-based distribution of program codes using a global placement graph that calculates optimal node assignments based on error probability and effect, allowing for precalculated relocation plans to prioritize safety and minimize computing power and energy consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If software applications are redundantly executed in different computing nodes to prevent errors, then system reliability is improved, but device complexity increases due to the need to manage multiple instances and their assignment to control units
Solution Approach 1:
The patent segments the software architecture by separating the execution environment (control units) from the software instances (applications). Multiple instances of the same application can be distributed across different control units, allowing redundancy without requiring the entire system to be redesigned for fault tolerance. This segmentation enables independent management of each instance while maintaining overall system reliability.
Solution Approach 2:
The patent implements a universal placement mechanism that can dynamically assign software instances to any available control unit based on current system state. The same application instance type can be executed on different control units depending on workload, availability, and error conditions. This multi-functional assignment strategy allows the system to adapt to various failure scenarios while maintaining consistent management procedures.
2Reliability
If the minimum number of redundant instances and hardware segregation requirements are enforced, then system safety is improved, but the difficulty of detecting and measuring optimal placement increases due to the multidimensional constraints
Solution Approach 1:
The patent transforms the complex multidimensional placement problem into a more manageable form by dynamically adjusting placement parameters based on current system conditions. Instead of solving a static optimization problem with fixed constraints, the system continuously evaluates and adjusts instance placements according to changing workload, error states, and resource availability. This parameter-based approach simplifies the measurement and detection of optimal configurations.
Solution Approach 2:
The patent implements dynamic placement strategies where the assignment of software instances to control units is not fixed but adapts in response to system conditions. When errors occur or resources become available, the system can dynamically relocate instances to maintain safety requirements while optimizing resource utilization. This dynamic approach reduces the complexity of finding optimal placements by allowing incremental adjustments rather than requiring complete re-optimization.
3Reliability
If dynamic reconfiguration is performed quickly in response to failures, then system reliability is improved, but energy consumption increases due to the computing power required for real-time optimization
Solution Approach 1:
The patent performs preliminary placement optimizations in advance, evaluating multiple potential failure scenarios and pre-determining optimal instance placements for various error conditions. By pre-calculating placement strategies, the system reduces the computational burden during actual failure events, requiring only simple activation of pre-planned configurations rather than complex real-time optimization. This significantly reduces energy consumption during critical reconfiguration events.
Solution Approach 2:
The patent implements a two-phase approach where comprehensive optimization is performed during low-stress periods (skipping detailed real-time calculation), and only essential placement decisions are made during actual failures. The system skips complex computational steps during critical moments by relying on pre-evaluated placement options, rushing through only the necessary activation and switching operations rather than performing full optimization cycles during energy-constrained failure responses.
Data Source
AI summary
The invention relates to a method for the dynamic, context-based distribution of program codes in a control system of a vehicle. The control system has a plurality of control apparatuses for executing the program codes. The program codes are assigned to the corresponding control apparatuses of the control system by a global placement graph. In doing so, the global placement graph is calculated in a computing unit that is located outside of the control system. The data of the global placement graph are transmitted to the control system. It is provided for the global placement graph to be calculated based on the current configuration as well as known reconfigurations, wherein the selection of the next node of the global placement graph to be calculated is calculated using the probability of occurrence of an error as well as the effect of the error.