Local Sensor Navigation Mapping Without External HD Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles rely on external high-precision maps, which can be costly to produce and maintain, and may not function effectively in environments with weak GPS signals or indoor settings, posing safety risks and inefficiencies.
Innovation Solution
The development of a system that generates local high-precision navigation maps in real-time using sensors like stereo cameras and LiDAR, integrating navigation features with 3D environment information to enable autonomous navigation without relying on external maps, utilizing machine learning techniques for feature detection and pose estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If autonomous vehicles rely on external high-precision maps, then navigation accuracy is improved, but system cost and complexity increase
Solution Approach 1:
The autonomous vehicle performs self-localization by detecting navigation features (lane markings, traffic signs, pedestrians) in its environment and using these features to determine its own position and orientation. This eliminates the need for external high-precision maps, allowing the system to serve itself for navigation while reducing system complexity and cost.
2Reliability
If external high-precision maps are used, then navigation reliability is improved, but adaptability to new environments deteriorates
Solution Approach 1:
The system pre-identifies and tracks navigation features in the environment before making navigation decisions. By continuously detecting and tracking features such as lane markings, traffic signs, and pedestrians in real-time, the system prepares navigation data in advance, enabling reliable and adaptable navigation in dynamic environments without requiring pre-existing external maps.
3Reliability
If computational resources focus on dynamic obstacle detection, then safety is improved, but local navigation capability deteriorates
Solution Approach 1:
The system combines dynamic obstacle detection with local navigation feature detection into a unified processing framework. The same sensors and computational resources used for detecting obstacles are also utilized to identify navigation features like lane markings and traffic signs, allowing both safety and navigation capabilities to be maintained simultaneously without requiring separate dedicated systems.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for robust and efficient autonomous navigation with high precision, reducing costs and ensuring safety by using locally detected features to create accurate maps, even in environments without reliable external data.
Implementation Method 1
stereo cameras that scan the roads along pre-planned routes and gather image data
Implementation Method 2
LiDAR, which stands for light detection and ranging, is an optical distance measurement device that uses the time of flight (TOF) of a light pulse to calculate the distance to an object
Data Source
AI summary
Local sensing based navigation maps can form a basis for autonomous navigation of a mobile platform. An example method includes obtaining real-time environment information that indicates an environment within a proximity of the mobile platform based on first sensor(s) carried by the mobile platform, detecting navigation features based on sensor data obtained from the first sensor(s) or second sensor(s) carried by the mobile platform, integrating information corresponding to the navigation features with the environment information to generate a local navigation map, and generating navigation command(s) for controlling a motion of the mobile platform based on the local navigation map.