Vehicle Path Prediction Using Road and Trajectory Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current advanced driver assist systems (ADAS) for vehicles face challenges in accurately predicting the near future path of a vehicle, particularly due to unpredictable behavior of other road users, and there is a need for improved prediction accuracy.
Innovation Solution
A method and system that collect and preprocess vehicle driving data and sensor data to create a representation of object data, including previous positions, headings, and velocities, which is then processed in a deep neural network to predict the near future path, utilizing a combination of convolutional and recurrent neural networks for enhanced accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional ADAS path prediction methods are used, then the system can provide basic path prediction, but the prediction accuracy is insufficient due to unpredictable behavior of road users
Solution Approach 1:
The system segments the path prediction task into multiple components: current state estimation, historical trajectory analysis, and future path prediction. By dividing the complex prediction problem into manageable segments, the system can apply specialized processing to each component, improving overall accuracy while maintaining robustness against unpredictable behaviors.
Solution Approach 2:
The system performs preliminary actions by collecting and analyzing historical trajectory data of road users before making path predictions. This advance preparation allows the system to establish baseline behaviors and detect anomalies, improving prediction accuracy while accounting for unpredictable deviations from normal patterns.
2Measurement precision
If more sensor data and processing are used to improve prediction accuracy, then the system can better predict paths, but the computational complexity and processing requirements increase
Solution Approach 1:
The system employs a unified deep neural network architecture that handles multiple functions: feature extraction from sensor data, historical trajectory analysis, and path prediction. This multi-functional approach improves prediction accuracy while avoiding the complexity of multiple separate processing systems by consolidating functions into a single versatile framework.
Solution Approach 2:
The system introduces an intermediary processing layer that transforms raw sensor data and historical trajectories into meaningful features before feeding them to the prediction model. This intermediary representation layer simplifies the computational burden on the neural network while preserving critical information for accurate path prediction.
3Measurement precision
If real-time processing of multiple data sources is performed, then the system can provide accurate path predictions, but the processing time and computational load increase
Solution Approach 1:
The system performs preliminary processing of sensor data and historical trajectories to extract essential features before the actual prediction step. This advance feature extraction reduces the computational load during real-time operation, maintaining high prediction accuracy while minimizing processing time and meeting real-time requirements.
Data Source
AI summary
A method and system for predicting a near future path for a vehicle. For predicting the near future path sensor data and vehicle driving data is collected. Road data is collected indicative of a roadway on the presently occupied road for the vehicle. The sensor data and the vehicle driving data is pre-processed to provide object data comprising a time series of previous positions, headings, and velocities of each of the objects relative the vehicle. The object data, the vehicle driving data, and the road data is processed in a deep neural network to predict the near future path for the vehicle. The invention also relates to a vehicle comprising the system.