Autonomous Driving Preprocessing Offload for Safety and Latency

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous driving systems face performance limitations due to cost, power consumption, and space constraints, leading to insufficient performance in autonomous movement, despite efforts to improve safety through hardware and software multiplexing.

Innovation Solution

An information processing method that offloads advanced preprocessing tasks from limited onboard computers to external servers, allowing for more accurate, faster, and less delayed processing, which improves the performance of autonomous movement by using advanced algorithms and resources not limited by cost or space constraints.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If hardware or software multiplexing is implemented to improve safety, then safety is improved, but performance in autonomous movement becomes insufficient

Engineering Contradiction:
ImprovesafetyVSAvoidperformance in autonomous movement
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the autonomous driving system into multiple independent processing units with different functions (detection, recognition, prediction, control). Each unit can be independently optimized for safety or performance based on its specific role, allowing the system to achieve both safety and performance simultaneously through functional decomposition rather than treating it as a monolithic system.

Inventive Principle:
Principle #1Segmentation

2Productivity

If more powerful computers are mounted on autonomous moving bodies to improve performance, then performance is improved, but cost, power consumption, and space increase

Engineering Contradiction:
Improveperformance in autonomous movementVSAvoidpower consumption
Core Design Contradiction:
ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent implements partial action by having different processing units operate at different levels of computational intensity based on their specific functions. Critical safety-related detection units use lower-power sensors and processors, while performance-critical recognition and prediction units use more powerful but selective processing. This allows the system to achieve high overall performance without all components consuming maximum power, resolving the contradiction between performance and power consumption.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4105907B1Information processing method, information processing system, and information processing device
Publication Date: 2024.12.04 PANASONIC INTELLECTUAL PROPERTY MANAGEMENT CO LTD
  • EP4105907B1 patent drawingFigure 1
  • EP4105907B1 patent drawingFigure 2
  • EP4105907B1 patent drawingFigure 3

AI summary

An information processing method is to be executed by a computer. The information processing method includes: receiving, from an autonomous moving body, a first processing result that is a result of first preprocessing of travel control processing in autonomous movement processing of the autonomous moving body, and sensing data received by the autonomous moving body (step S11); executing second preprocessing based on the sensing data to receive a second processing result, the second preprocessing being more advanced than the first preprocessing (step S12); determining a difference between the first processing result and the second processing result (step S15); and outputting, to the autonomous moving body based on the difference determined, a change request to change the first processing result to the third processing result (step S17), the third processing result being received based on at least one of the first processing result or the second processing result.