Mapless Traveling Plan Generation for Unmapped Autonomous Driving

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Autonomous driving systems face challenges in executing safe navigation in regions where detailed autonomous driving map information is not available, as they lack corresponding data for all vehicle traveling regions.

Innovation Solution

A vehicle control device that includes a map non-correspondence region recognition unit, traveling environment recognition unit, traveling environment classification database, and a vehicle control unit, which generates a mapless traveling plan based on past traveling plans and their associated results to enable safe autonomous driving in unmapped areas by determining the most suitable environment classification and risk index.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If autonomous driving map information is used for navigation, then driving accuracy in mapped regions is improved, but autonomous driving cannot be executed in unmapped regions

Engineering Contradiction:

Improvedriving accuracyVSAvoidautonomous driving capability in unmapped regions

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent introduces mapless traveling plan generation as an intermediary mechanism that bridges the gap between mapped and unmapped regions. When the vehicle enters a map non-correspondence region, the system generates alternative traveling plans based on sensor data and historical information, allowing continuous autonomous operation without complete reliance on pre-mapped data

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system performs preliminary classification of traveling environments using historical data before encountering unmapped regions. By pre-categorizing environmental characteristics and storing them as historical information, the system prepares advance guidance strategies that can be quickly deployed when map data becomes unavailable

Inventive Principle:

Principle #10Preliminary action

2Adaptability or versatility

If detailed autonomous driving map information is collected for all regions, then coverage is improved, but system complexity and data requirements increase

Engineering Contradiction:

Improvemap coverageVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the operating environment into mapped regions and map non-correspondence regions, allowing the system to use different navigation strategies for each. This segmentation avoids the need to collect and process detailed map data for all possible regions, reducing overall system complexity while maintaining capability in critical areas

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of maintaining expensive, comprehensive map data for all regions, the system uses inexpensive, locally-generated sensor data and historical traveling information as temporary substitutes when map data is unavailable. This approach replaces the need for extensive permanent map coverage with transient, on-demand environmental understanding

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

3Adaptability or versatility

If mapless traveling plan is generated using historical data, then autonomous driving in unmapped regions is enabled, but reliability depends on data availability

Engineering Contradiction:

Improveautonomous driving in unmapped regionsVSAvoidsystem reliability

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system continuously monitors the correspondence between current location and map data, providing feedback that triggers transitions between map-based and mapless navigation modes. This feedback mechanism ensures reliable operation by maintaining map-based navigation when available and smoothly transitioning to historical data-based navigation only when necessary

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system accumulates historical traveling environment data and classified information in advance as a cushion against future map data unavailability. By preparing this reservoir of historical information beforehand, the system ensures reliable autonomous operation in unmapped regions when encountered, reducing dependence on real-time data availability

Inventive Principle:

Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentUS10976739B2Vehicle control device

Publication Date: 2021.04.13 TOYOTA JIDOSHA KK

AI summary

A vehicle control device includes a traveling environment classification database configured to record traveling environment classification data in which at least one past traveling plan for autonomous driving and a traveling result of the past traveling plan are associated with each other for each preset traveling environment classification, a traveling environment classification determination unit configured to determine whether or not the traveling environment classification corresponding to a traveling environment of a host vehicle is present in the traveling environment classification database when the host vehicle travels in a map non-correspondence region during the autonomous driving, and a traveling plan generation unit configured to generate a mapless traveling plan as a traveling plan of the host vehicle in the map non-correspondence region when the traveling environment classification corresponding to the traveling environment of the host vehicle is present in the traveling environment classification database.