Dual-SoC Autonomous Driving Control Tuned to Driver Habits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving systems do not effectively reflect a driver's driving characteristics and habits, leading to decreased reliability and satisfaction.
Innovation Solution
A vehicle system on chip (SoC) that includes a first SoC to determine manual driving characteristics and a second SoC to correct autonomous driving settings based on these characteristics, ensuring the autonomous driving function aligns with the driver's habits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous driving control is performed based on standard setting values from neural network operations, then the autonomous driving function can be implemented, but the system does not reflect driver's driving characteristics and habit, causing decreased reliability
Solution Approach 1:
The system performs preliminary action by collecting and analyzing driver's manual driving characteristics before implementing autonomous driving. The first SoC processes sensor information and manual operation information to establish a baseline of driver behavior patterns, which is then used to adjust autonomous driving setting values, ensuring the system adapts to individual driver preferences rather than using generic algorithms
Solution Approach 2:
The system implements feedback by continuously comparing the driver's manual driving characteristics with the autonomous driving control outputs. The second SoC receives feedback from the first SoC about driver preferences and adjusts the setting values accordingly, creating a closed-loop system that learns from and adapts to driver behavior, thereby improving reliability through personalized adaptation
2Adaptability or versatility
If a single SoC is used for autonomous driving control, then the device complexity is reduced, but the system cannot simultaneously capture manual driving characteristics and correct autonomous driving settings in real-time
Solution Approach 1:
The system applies segmentation by dividing the autonomous driving control function into two distinct SoCs with specialized roles. The first SoC is dedicated to processing sensor information and analyzing manual driving characteristics, while the second SoC focuses on generating and correcting autonomous driving setting values. This functional segmentation allows each unit to optimize its specific task while collectively achieving real-time adaptation without excessive complexity
3Adaptability or versatility
If autonomous driving setting values are fixed based on neural network operations, then the control algorithm is simple, but the system cannot account for individual driver preferences and circumstances
Solution Approach 1:
The system performs preliminary action by pre-processing and analyzing driver behavior patterns during manual driving phases. The first SoC captures and stores driver preferences and characteristics before autonomous driving engagement, creating a ready-to-use profile that the second SoC can quickly reference and apply, thus maintaining processing speed while achieving personalized adaptation
Solution Approach 2:
The system implements feedback by continuously monitoring the effectiveness of autonomous driving setting values and comparing them with driver preferences. The second SoC receives real-time feedback from the first SoC about driver reactions and adjusts setting values dynamically, creating an adaptive system that balances processing efficiency with personalized customization based on actual driver responses
Data Source
AI summary
A system on chip (SoC), an autonomous driving system including the same, and an operating method thereof is provided. The autonomous driving system includes: a first SoC configured to receive, in a manual driving mode of a vehicle, pieces of sensor information from sensors mounted on the vehicle and generate first information indicating manual driving characteristics of a driver by processing the pieces of sensor information and manual operation information generated when the driver operates the vehicle; and a second SoC configured to receive the pieces of sensor information, generate, in the manual driving mode, second information including one or more setting values for controlling the vehicle according to an autonomous driving function, and correct in real-time the one or more setting values related to the autonomous driving function by comparing, in the manual driving mode, the first information to the second information.