Autonomous Driving Path Planning Under Low Log Storage Capacity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In-vehicle storage devices have limited capacity, posing a risk of insufficient storage for data logs related to autonomous driving control, which can lead to data loss if not managed effectively.
Innovation Solution
The system acquires and monitors the remaining storage capacity, generating recognition data with reduced accuracy when capacity is low to minimize data volume, allowing for continued storage of essential logs and ensuring safety margins in path planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high accuracy recognition data is stored in the data log, then the verification accuracy of autonomous driving control is improved, but the storage capacity is consumed faster
Solution Approach 1:
The patent applies local quality by differentiating the accuracy requirements of different data types in the data log. High accuracy is maintained for critical path plan data that directly affects safety, while reduced accuracy is acceptable for peripheral recognition data. This selective quality approach allows verification of essential autonomous driving functions while consuming less storage capacity.
Solution Approach 2:
The patent changes the accuracy parameter of recognition data based on storage conditions. When storage capacity is sufficient, full accuracy data is stored. When storage capacity becomes limited, the system transitions to storing reduced accuracy data, thereby extending the storage duration and preventing data loss while maintaining adequate verification capability.
2Quantity of substance
If the in-vehicle storage device capacity is increased, then more data logs can be stored, but the device complexity and cost increase
Solution Approach 1:
The patent implements a dynamic data quality adjustment mechanism that adapts to available storage capacity. The system continuously monitors storage levels and dynamically changes the accuracy level of stored recognition data accordingly. This dynamic approach allows the system to maximize storage utilization without requiring additional hardware capacity, thereby avoiding increased device complexity and cost.
3Quantity of substance
If reduced accuracy recognition data is used, then the data volume is reduced and storage capacity is preserved, but the path plan generation accuracy decreases
Solution Approach 1:
The patent applies local quality by ensuring that critical path plan generation data maintains high accuracy while peripheral recognition data can use reduced accuracy. The system selectively applies full accuracy processing to data elements that directly impact safety-critical path planning decisions, while allowing compression for less critical data, thus balancing storage efficiency with path plan generation accuracy.
Solution Approach 2:
The patent applies partial action by selectively reducing accuracy only for specific types of recognition data that are less critical to path plan generation. Essential data elements required for safe autonomous driving control maintain full accuracy, while non-critical elements use reduced accuracy representations. This selective approach preserves adequate path plan generation accuracy while achieving data volume reduction.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to an autonomous driving system (100) mounted on a vehicle (1). The autonomous driving system (100) comprises a storage device and one or more processors (110). The one or more processors (110) are configured to execute acquiring recognition data by recognizing a situation around the vehicle (1), generating a path plan (PLN) for the vehicle (1) based on the recognition data, performing autonomous driving control of the vehicle (1) in accordance with the path plan (PLN), and storing a data log related to the autonomous driving control in the storage device. The data log includes a log of data used for generating the path plan (PLN). Generating the path plan (PLN) includes acquiring a remaining capacity of the storage device, generating target data being the recognition data with reduced accuracy, and generating the path plan (PLN) that can be generated using the target data when the remaining capacity is equal to or less than a predetermined amount.