Autonomous Driving Route Planning Across Different Sensor Specs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving vehicles require significant development effort to adapt route plans due to differences in sensor configurations, including number, type, and detection performance, between vehicles.
Innovation Solution
A method involving training a machine learning model using a first autonomous driving vehicle with superior sensor specifications and then applying this model to a second vehicle with inferior sensor specifications, allowing for equivalent route output without separate re-adaptation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If route plans are separately re-adapted for each autonomous driving vehicle with different sensor configurations, then the route plan accuracy for each vehicle is improved, but the development time and man-hours increase significantly
Solution Approach 1:
The patent creates a unified machine learning model that can universally process detection results from different sensor configurations (cameras, LiDAR, radar) and output appropriate route plans. The model is trained on diverse sensor data during the learning phase, enabling it to adapt to various sensor types and configurations without requiring separate development for each vehicle setup, thus achieving both route plan accuracy and reduced development time
2Adaptability or versatility
If a machine learning model is trained on data from a vehicle with superior sensor specifications, then the model can be applied to vehicles with inferior sensor specifications, but the detection performance may be compromised
Solution Approach 1:
The patent employs parameter changes by adjusting the learning conditions and data processing parameters based on the specific sensor configuration of each vehicle. The machine learning model is trained with appropriate parameter settings that account for differences in sensor quality, types, and configurations, allowing the same model architecture to adapt to various sensor specifications while maintaining optimal detection performance for each vehicle type
Data Source
AI summary
A method for developing an autonomous driving vehicle including an external sensor including a number, a type, and a detection performance as a sensor specification, wherein the autonomous driving vehicle includes a machine learning model to configured to output a route with respect to at least a recognition process and a planning process to plan the route among the recognition process, the planning process, and a vehicle control process, the method comprising: a first step of training the machine learning model by adapting at least the recognition process and the planning process using a first autonomous driving vehicle including the external sensor having a first sensor specification; and a second step of applying the machine learning model trained in the first step to a second autonomous driving vehicle including the external sensor having a second sensor specification that is inferior to the first sensor specification.