Autonomous Driving Log Storage Control by Memory-Based Mileage Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The limited memory capacity in vehicles performing autonomous driving using machine learning models leads to a situation where log data generated during autonomous driving control cannot be stored due to filled memory space, necessitating a technique to prevent this tight memory situation.

Innovation Solution

The autonomous driving system predicts the vehicle's mileage based on the available free space in the memory device, setting the mileage to a shorter distance as the free space decreases, thereby avoiding the situation where memory space becomes too tight.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If log data is continuously stored in the memory device during autonomous driving control, then the verification capability is improved, but the memory device becomes full and cannot store new log data

Engineering Contradiction:
Improveverification capabilityVSAvoidmemory capacity
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The system predicts the mileage before it is exhausted based on the current free space in the memory device, and issues warnings or restrictions in advance to prevent the memory from becoming full. This allows the system to take preventive action before the problem occurs, ensuring continuous log data storage capability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system continuously monitors the free space in the memory device and uses this feedback to predict remaining mileage. When the predicted mileage becomes short, the system provides feedback through warnings or restrictions to the user, enabling dynamic adjustment of usage behavior to avoid memory exhaustion.

Inventive Principle:
Principle #23Feedback

2Productivity

If the vehicle operates for longer mileage, then the productivity is improved, but the log data fills the memory device and prevents new data storage

Engineering Contradiction:
Improveoperational mileageVSAvoidlog data storage capability
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The system predicts how much mileage can be operated before the memory becomes full, and provides this information in advance to the user. This allows users to plan their usage and transfer or delete log data before the memory is exhausted, maintaining continuous operational capability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The predicted mileage is not a fixed value but dynamically changes based on the current free space in the memory device. As log data is added or removed, the predicted mileage updates accordingly, providing real-time guidance for maintaining both productivity and storage capability.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4498340A1Autonomous driving system
Publication Date: 2025.01.29 TOYOTA JIDOSHA KK
  • EP4498340A1 patent drawingFigure 1
  • EP4498340A1 patent drawingFigure 2
  • EP4498340A1 patent drawingFigure 3

AI summary

Processing for autonomous driving control of a vehicle (1), processing to store log data related to the autonomous driving control in one or more memory devices (120) during an execution of the autonomous driving control, and processing to predict a mileage of the vehicle is executed. In the processing to predict the mileage, a free space of the one or more memory devices in which log data is stored is acquired. In addition, the mileage is set according to the acquired free space. The mileage is set to be shorter as the free space is smaller.