Automated Driving Path Planner for Dynamic Object Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated driving systems lack the ability to autonomously navigate vehicles based on real-time environmental features, relying heavily on pre-mapped routes and failing to adapt to dynamic objects of interest effectively.
Innovation Solution
An automated driving system equipped with sensors and a computing device that determines potential vehicle paths, classifies objects of interest, and selects a preferred path based on their properties, sending commands to control the vehicle's systems to follow the chosen path, integrating data from various sensors like LIDAR, GNSS, and dead-reckoning systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the automated driving system uses pre-mapped routes for navigation, then the navigation route is well-defined and easy to follow, but the system cannot adapt to dynamic objects of interest in the environment
Solution Approach 1:
The system segments the navigation task into distinct modules: a route determination module that handles pre-mapped routes and a path selection module that handles dynamic path adjustments based on detected objects. This segmentation allows the system to maintain simple pre-mapped route navigation while adding adaptive path selection capabilities without overwhelming system complexity.
Solution Approach 2:
The path selection module dynamically adjusts the vehicle path based on real-time detection of objects of interest. The system transitions from static pre-mapped routes to dynamic path adjustment by evaluating object properties (such as whether objects are vehicles, pedestrians, or obstacles) and selectively applying path modifications only when necessary, maintaining adaptability while controlling complexity.
2Productivity
If the system considers all objects in the environment, then the navigation is highly adaptive, but the processing time and computational load increase significantly
Solution Approach 1:
The system extracts only the relevant features from detected objects that are necessary for path selection, such as object type, position, and potential hazard level. By taking out only the essential information rather than processing all object properties, the system maintains high navigation adaptability while reducing computational load and path selection time.
Solution Approach 2:
The system applies partial action by selectively adjusting the vehicle path only when objects of interest are detected that require avoidance or special attention. Instead of continuously modifying the path for all detected objects, the system intervenes only when necessary, improving navigation efficiency without excessive processing time.
3Reliability
If the automated driving system relies heavily on pre-mapped routes, then the navigation is simple and fast, but the system fails to adapt to real-time environmental features and objects of interest
Solution Approach 1:
The system merges pre-mapped route information with real-time sensor data from LIDAR, GNSS, and dead-reckoning systems. The route determination module combines these data sources to establish the navigation route, while the path selection module integrates object detection data to adjust the path when objects of interest are present. This merging enhances navigation safety by combining multiple information sources while managing sensor integration complexity through modular architecture.
Solution Approach 2:
The path selection module acts as an intermediary between the pre-mapped route determination and the vehicle control systems. It receives the navigation route from the route determination module, processes real-time object detection data, and generates adjusted paths when objects of interest are detected. This intermediary role allows the system to maintain reliability by preserving the pre-mapped route foundation while adapting to real-time environmental features.
Data Source
AI summary
A system, device, and methods of automated driving are disclosed. One example method includes determining one or more potential vehicle paths based on information specific to the environment surrounding a vehicle and receiving an indication classifying one or more objects proximate to the one or more potential vehicle paths as an object of interest based on input received from one or more sensors disposed on the vehicle. The method further includes selecting a preferred vehicle path from the one or more potential vehicle paths based on properties of the one or more objects of interest and sending a command, to one or more vehicle systems, to control the vehicle to follow the preferred vehicle path.