Autonomous Vehicle Task Scheduling for Reboot-Free Mode Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous vehicle control systems require rebooting whenever operation modes change between driving and parking, causing time delays and increasing hardware complexity and cost.
Innovation Solution
A control apparatus for autonomous vehicles that uses a single hardware processor to manage both driving and parking modes by scheduling tasks and adjusting execution times based on the current operation mode using a scheduler.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the control apparatus performs rebooting whenever the operation mode is changed, then the system can load the correct files for the new mode, but a time delay occurs which is unacceptable in real-time autonomous driving environments
Solution Approach 1:
The scheduler pre-loads files for different operation modes into the RAM before mode switching occurs. When a mode change is requested, the corresponding files are already available in memory, eliminating the need for rebooting and file loading delays. This preliminary preparation ensures both fast response and correct file loading.
Solution Approach 2:
The system dynamically adjusts task execution by scheduling different tasks for different operation modes without requiring a reboot. The scheduler dynamically allocates CPU time slices to appropriate tasks based on the current mode, allowing seamless transitions between driving and parking modes while maintaining real-time performance.
2Reliability
If separate hardware processors are used for driving and parking operations, then each mode can be handled independently, but the structural complexity and manufacturing cost increase
Solution Approach 1:
A single hardware processor is designed to perform multiple functions by executing different tasks for different operation modes. The scheduler manages this single processor to handle both driving and parking operations by allocating time slices to mode-specific tasks, eliminating the need for separate processors while maintaining independent handling of each mode.
Solution Approach 2:
The operation modes are segmented into distinct tasks that are scheduled separately by the scheduler. Each task is designed to handle specific mode requirements (driving tasks or parking tasks), allowing the single processor to efficiently manage different modes through task segmentation rather than hardware segmentation.
Data Source
AI summary
A control apparatus and method for an autonomous vehicle includes a plurality of cores, and a scheduler that schedules a plurality of tasks corresponding to each operation mode of the autonomous vehicle for each of the cores, and adjusts an execution time of a task in the each core based on a current operation mode of the autonomous vehicle.


