Autonomous Driving Trajectory Mapping for No-Lane Sections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving systems face challenges in navigating no-lane sections without recognizable lane markings, such as near intersections or toll houses, as they rely on detectable road features for route planning and control.
Innovation Solution
A vehicle travel assistance method that collects and analyzes probe information from multiple vehicles to generate statistical trajectory information for no-lane sections, which is then used to create route information and enable autonomous driving by formulating a travel plan based on this data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous driving systems rely on detectable road features (lane markings) for route planning and control, then the system can maintain stable lane-keeping assistance, but it cannot navigate no-lane sections where lane markings are absent or unrecognizable
Solution Approach 1:
The system performs preliminary actions by collecting probe information from multiple vehicles and pre-calculating statistical trajectory information for no-lane sections before autonomous driving is needed. This pre-computed route information is stored and readily available when the vehicle encounters a no-lane section, enabling seamless navigation without real-time detection challenges
Solution Approach 2:
The patent introduces statistical trajectory information as an intermediary element that bridges the gap between detectable lane markings and unrecognizable no-lane sections. This intermediary data structure, derived from aggregated probe information, provides a virtual reference framework that enables continuous autonomous driving control even when physical lane markings are absent
2Loss of information
If the system uses probe information from multiple vehicles to generate statistical trajectory information, then route information for no-lane sections can be obtained, but data processing complexity increases
Solution Approach 1:
The system merges probe information from multiple vehicles traveling through the same no-lane section to generate statistical trajectory information. By combining data from multiple sources and applying statistical processing, the system creates a robust route representation that compensates for individual vehicle data limitations while managing processing complexity through aggregation
Solution Approach 2:
The patent creates a virtual copy of the physical road geometry by generating statistical trajectory information that replicates the expected path through no-lane sections. This copied geometric information, derived from aggregated probe data, serves as a substitute for actual lane markings, enabling autonomous driving without requiring complex real-time sensing in featureless areas
Data Source
AI summary
Autonomous driving is possible even in a no-lane section. Probe information to be transmitted from a plurality of vehicles is stored. Vehicle travel trajectory information in a no-lane section is extracted from the stored probe information, the vehicle travel trajectory information is extracted per combination of an entry point and an exit point for the no-lane section, the extracted vehicle travel trajectory information per combination of the entry point and exit point for the no-lane section is sorted into a plurality of categories according to a predetermined criterion, and statistical trajectory information is calculated by statistical processing of the travel trajectory information for each of the plurality of categories. The calculated statistical trajectory information is transmitted to the plurality of vehicles as vehicle route information in the no-lane section.