Hybrid Path Planning for Autonomous Driving Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vehicles with autonomous driving capabilities face challenges in accurately processing and responding to road elements in real-time, necessitating improved information processing to enhance decision-making.
Innovation Solution
A method utilizing a combination of pixel-based and object-based path planning outputs generated by machine learning processes, integrated through a group of artificial intelligence agents, to enhance the accuracy and reliability of driving decisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If only object-based path planning is used, then processing speed is improved, but measurement precision of road elements deteriorates
Solution Approach 1:
The patent combines object-based path planning with pixel-based path planning to create a hybrid approach. The system integrates the speed advantages of object-based planning with the precision advantages of pixel-based planning, using both methods to generate path planning outputs that are then fused to make driving decisions.
Solution Approach 2:
The patent segments the path planning process into two distinct approaches: object-based path planning for speed and pixel-based path planning for precision. Each approach processes information differently, and both segments contribute to the final decision-making process through their respective outputs.
2Measurement precision
If only pixel-based path planning is used, then measurement precision of road elements is improved, but processing speed deteriorates
Solution Approach 1:
The patent merges pixel-based path planning with object-based path planning to balance precision and speed. The system uses both planning approaches simultaneously, leveraging the high precision of pixel-based methods while compensating for their computational intensity through the faster object-based approach.
Solution Approach 2:
The patent applies partial action by using pixel-based path planning selectively for scenarios requiring high precision while relying more on object-based planning for standard situations. This allows the system to achieve necessary precision without always incurring the full computational cost of pixel-based processing.
3Device complexity
If a single path planning output is used, then device complexity is reduced, but reliability of driving decisions deteriorates
Solution Approach 1:
The patent combines multiple path planning outputs from both object-based and pixel-based approaches to enhance reliability. By integrating these diverse outputs, the system achieves more robust and reliable driving decisions while managing complexity through a unified processing framework.
Solution Approach 2:
The patent implements feedback mechanisms where the system evaluates the confidence levels of different path planning outputs and adjusts its decision-making accordingly. This feedback loop allows the system to rely more on reliable outputs and compensate for uncertainties, improving overall decision reliability.
4Reliability
If multiple path planning outputs are fused, then reliability of driving decisions is improved, but device complexity increases
Solution Approach 1:
The patent introduces an intermediary processing layer that fuses path planning outputs from multiple sources. This intermediary component manages the complexity of integrating multiple outputs by providing a structured approach to combining information, thereby improving reliability without proportionally increasing overall system complexity.
Solution Approach 2:
The patent creates a universal processing framework that handles multiple path planning outputs through a common decision-making architecture. This multi-functional system can process different types of path planning data using the same core mechanisms, reducing the complexity increase that would otherwise result from handling multiple specialized processing paths.
Data Source
AI summary
A method of a pixel based with object based decision making for driving, the method includes receiving, at a first machine learning process of an artificial intelligence agent, a sensed information unit; receiving, at a second machine learning process of the artificial intelligence agent, object descriptive information regarding an object captured in the sensed information unit; generating, by the first machine learning process, a pixel-based path planning output related to a suggested pixel-based path segment of a vehicle; generating, by the second machine learning process, an object-based path planning output related to a suggested object-based path segment of the vehicle; and generating, by at least in part processing the pixel-based path planning output in correspondence with the object-based path planning output, a driving related output with respect to the vehicle.