Vehicle Telemetry Rasterization for Sparse Roadway Map Content
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in creating map content for autonomous vehicles from telemetry data due to temporal sparsity, spatial non-uniformity, noise, and the vast size of the data, which makes it difficult to process and infer roadway geometry from individual vehicle paths.
Innovation Solution
A system utilizing central computers to align telemetry data points with a road network, calculate splines, interpolate data points, and rasterize them into discrete pixels, then determine and render probability density images based on vehicle trajectories, incorporating GPS boundary errors and filtering parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If telemetry data is aggregated from numerous vehicles to create map content, then the quantity of data increases, but the processing difficulty and computational load increase significantly
Solution Approach 1:
The patent segments the continuous telemetry data into discrete trajectory points and further divides the processing into modular steps: data collection, trajectory generation, rasterization, and probability density calculation. This segmentation allows parallel processing and reduces the complexity of handling vast amounts of data as a single monolithic task.
Solution Approach 2:
The patent transforms trajectory data from continuous spatial coordinates into a discrete grid-based probability density map. By converting individual vehicle trajectories into aggregated probability distributions on a grid, the system reduces the dimensionality and complexity of processing while preserving essential spatial patterns.
2Productivity
If telemetry data is collected at sparse time intervals, then data collection efficiency improves, but the temporal precision and completeness of trajectory information deteriorates
Solution Approach 1:
The system performs preliminary interpolation to generate equidistant trajectory points between sparsely sampled telemetry data points. This preliminary action reconstructs the missing temporal information by calculating intermediate positions based on the vehicle's trajectory, thereby recovering temporal precision without requiring continuous data collection.
Solution Approach 2:
The patent creates a simplified copy of the trajectory as a spline curve that passes through the sparse data points. This spline representation captures the essential temporal-spatial relationship without requiring the original high-frequency data, allowing efficient processing while maintaining acceptable temporal precision.
3Measurement precision
If individual vehicle trajectories are processed separately to maintain data integrity, then measurement accuracy is preserved, but the ability to infer roadway geometry from aggregated data is reduced
Solution Approach 1:
The patent merges multiple individual trajectory lines into a single aggregated probability density map. By overlaying and combining trajectories from numerous vehicles on the same roadway, the system infers the underlying road geometry and characteristics that are not apparent in individual trajectories, thereby recovering lost information while preserving individual measurement accuracy.
Solution Approach 2:
The probability density function serves as an intermediary that bridges individual trajectory data and roadway geometry inference. It aggregates individual trajectories while preserving their statistical properties, allowing the system to infer road characteristics without losing the accuracy of individual measurements.
4Quantity of substance
If telemetry data includes noise from vehicles traveling through unmapped areas, then data completeness improves, but the quality and reliability of map content deteriorates
Solution Approach 1:
The system extracts and separates valid trajectory segments from noise by identifying portions of trajectories that align with the road network. It removes or excludes data from vehicles traveling through unmapped or irrelevant areas, thereby maintaining data completeness for relevant regions while improving map content reliability by eliminating noisy contributions.
Data Source
AI summary
A system for rasterizing telemetry data collected from a plurality of vehicles into map content includes one or more central computers in wireless communication with the plurality of vehicles that are each situated within a predefined geofenced area. The one or more central computers receive the telemetry data from each of the plurality of vehicles. The one or more central computers determine a two-dimensional probability distribution indicating a probability density of the plurality of vehicles over a predefined period of time along a particular road segment located within the predefined geofenced area.