Object Trajectory Prediction Using Map and Context Cues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately predicting the trajectories of objects in their environment, leading to potential collisions and hazardous situations due to the limitations of existing perception systems in combining detailed map information with contextual data.
Innovation Solution
A computer-implemented method and system that identifies objects in a vehicle's environment, generates possible actions and trajectories based on map information and contextual data, determines likelihood values for these trajectories, and maneuvers the vehicle to avoid predicted trajectories by selecting the most likely ones, using thresholds and situational relationships to filter and prioritize actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the perception system uses basic sensor data without integrating detailed map information and contextual data, then the system complexity is low, but the trajectory prediction accuracy is insufficient leading to potential collisions
Solution Approach 1:
The patent combines multiple data sources including sensor data, detailed map information, and contextual data into a unified perception system. This integration allows the system to generate multiple possible trajectories by merging information from cameras, sensors, map databases, and contextual analysis, thereby improving prediction accuracy while managing complexity through systematic data fusion
Solution Approach 2:
The perception system segments the trajectory prediction task into distinct components: generating multiple possible trajectories, assigning probability values to each trajectory, and selecting the most likely trajectory. This segmentation allows each component to be processed independently and improves overall prediction accuracy without overwhelming system complexity
2Reliability
If the system generates and evaluates multiple possible trajectories with probability values, then the safety and collision avoidance improve, but the computational time and processing complexity increase
Solution Approach 1:
The system performs preliminary actions by pre-generating multiple possible trajectories and assigning probability values to each before making a final decision. This allows the autonomous vehicle to evaluate various potential paths in advance, improving safety through comprehensive analysis while managing computational time by preparing trajectory options proactively rather than reactively
Solution Approach 2:
The perception system generates an excessive number of possible trajectories beyond what a single deterministic path would provide. By creating multiple hypothetical trajectories with varying probability values, the system ensures thorough safety evaluation, accepting increased computational time as a necessary trade-off for enhanced reliability and collision avoidance
3Measurement precision
If the system integrates detailed map information with real-time sensor data and contextual information, then the navigation accuracy improves, but the data processing complexity and energy consumption increase
Solution Approach 1:
The perception system is designed with multi-functionality to handle diverse data types including map information, sensor data, and contextual information through a unified processing framework. This universal approach improves navigation accuracy by integrating multiple information sources while managing energy consumption through efficient data fusion algorithms that can adaptively prioritize processing based on operational context
Data Source
AI summary
Aspects of the disclosure relate to detecting and responding to objects in a vehicle's environment. For example, an object may be identified in a vehicle's environment, the object having a heading and location. A set of possible actions for the object may be generated using map information describing the vehicle's environment and the heading and location of the object. A set of possible future trajectories of the object may be generated based on the set of possible actions. A likelihood value of each trajectory of the set of possible future trajectories may be determined based on contextual information including a status of the detected object. A final future trajectory is determined based on the determined likelihood value for each trajectory of the set of possible future trajectories. The vehicle is then maneuvered in order to avoid the final future trajectory and the object.