Radar Corridor Prediction Using Segmented Detection Point Encoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing driving assistance systems, particularly those utilizing radar sensors, face challenges in processing detection points due to their randomness, which hinders effective prediction and control of vehicle trajectories.

Innovation Solution

A computer-implemented method that encodes detection points from radar sensors into latent features, determines fixed features for segments, and calculates spatial correlations to indicate driving corridor characteristics, using machine learning modules for segment and correlation encoding, enabling improved processing and utilization of radar data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If radar detection points are used directly for prediction, then the system can process raw sensor data, but the randomness of radar data reduces prediction accuracy

Engineering Contradiction:
Improveprediction accuracyVSAvoiddata consistency
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The region of interest is divided into multiple segments, and detection points are processed segment by segment. This segmentation allows the system to handle radar data in organized units, applying encoding and aggregation operations to reduce randomness while preserving spatial information for accurate prediction.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediate processing steps including encoding detection points into latent features, determining fixed features for segments, and encoding fixed features to obtain spatial correlations. These intermediary representations transform raw random radar data into structured information suitable for prediction.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of information

If all detection points are processed individually, then detailed object information is preserved, but the system complexity increases

Engineering Contradiction:
Improveobject information retentionVSAvoidprocessing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

Detection points within each segment are merged through encoding into latent features, and fixed features are determined by aggregating latent features. This merging reduces the number of individual processing units while retaining essential object information through compressed representations.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent transforms detection points from spatial coordinates into latent feature space through encoding, and then aggregates these features to determine fixed features. This dimensional transformation allows information retention while simplifying subsequent processing operations.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Measurement precision

If the region of interest is divided into many segments, then spatial correlation accuracy improves, but the processing time increases

Engineering Contradiction:
Improvespatial correlation accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The region of interest is divided into segments to capture spatial correlations at an appropriate granularity. This segmentation strikes a balance between achieving sufficient spatial accuracy and maintaining reasonable processing throughput by avoiding excessive segmentation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent processes segments in a continuous pipeline, encoding detection points and determining fixed features for each segment sequentially. This continuous processing minimizes idle time and maintains efficient data flow through the system while achieving accurate spatial correlations.

Inventive Principle:
Principle #20Continuity of useful action

Data Source

PatentEP4306988A1Integration of radar detection for prediction
Publication Date: 2024.01.17 APTIV TECHNOLOGIES AG
  • EP4306988A1 patent drawingFigure 1
  • EP4306988A1 patent drawingFigure 2
  • EP4306988A1 patent drawingFigure 3

AI summary

The subject disclosure pertains to advanced driving assistance systems and autonomous driving systems for vehicles. For example, the subject disclosure provides a computer-implemented method for a driving assistance system of a vehicle. The method comprises determining a plurality of detection points obtained by a sensor, the detection points being associated with a plurality of segments of a region of interest for the vehicle and representing objects in the plurality of segments. The method further comprises, for each segment of the plurality of segments, encoding the plurality of detection points associated to said segment into a latent feature of said detection point, and determining a fixed feature for said segment based on the latent features of the plurality of detection points. The method further comprises encoding the fixed features of the plurality of segments to obtain a spatial correlation of static detection points among the segments, the spatial correlation of static detection points indicating a characteristic of a driving corridor for the vehicle. The spatial correlation of static detection points is then provided to the driving assistance system.