Radar Point Cloud Compensation for Multipath Reflection Errors

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous vehicles face challenges in navigating due to the inability to accurately differentiate between single path and multipath data in point clouds, leading to incorrect object location determination and degraded data resolution.

Innovation Solution

A computing device equipped with a machine learning algorithm, such as a graphical neural network, is used to differentiate between single path and multipath data in point clouds, transforming indirect data representations into direct object location data, thereby generating corrected point cloud data for improved navigation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multipath data is discarded to avoid errors, then reliability of object location determination is improved, but loss of information increases and productivity decreases

Engineering Contradiction:
Improveobject location determination accuracyVSAvoidpoint cloud data utility
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent transforms multipath reflection data, which was previously considered harmful or useless, into beneficial information by using machine learning algorithms to distinguish valid multipath reflections from erroneous ones. The system learns to identify patterns in multipath data that correspond to actual objects, converting what was discarded waste into useful navigation information.

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The patent changes the parameter of data interpretation by applying machine learning models that alter how multipath data is processed. Instead of simple geometric assumptions, the system uses learned parameters and patterns to correctly interpret multipath reflections, transforming the data from unusable to useful form.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If machine learning algorithms are applied to differentiate single path and multipath data, then measurement precision of object location is improved, but device complexity increases

Engineering Contradiction:
Improveobject location accuracyVSAvoidprocessing system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces a machine learning model as an intermediary between raw radar data and object location determination. This intermediary layer processes the complex multipath data using learned patterns, acting as a mediator that translates raw sensor data into accurate location information without requiring complex geometric calculations or manual intervention.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical or geometric methods of point cloud interpretation with machine learning-based processing. Instead of using complex geometric algorithms to trace signal paths, the system uses data-driven models that have learned the patterns of multipath reflections, substituting computational mechanics with intelligent processing.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP4312054A1Radar point cloud multipath reflection compensation
Publication Date: 2024.01.31 GM CRUISE HOLDINGS LLC
  • EP4312054A1 patent drawingFigure 1A
  • EP4312054A1 patent drawingFigure 1B
  • EP4312054A1 patent drawingFigure 2A

AI summary

A method and apparatus for implementing a method are disclosed. The method includes providing point cloud data to a machine learning algorithm, the point cloud data detected in the vicinity of an autonomous vehicle. The method further includes differentiating, via the machine learning algorithm, in the point cloud, data directly representing a location of a first object and data indirectly representing a location of a second object. The method includes transforming the data indirectly representing the location of the second object into data directly representing the location of the second object and generating corrected point cloud data based on the data directly representing the location of the first object and the data directly representing the location of the second object. The method includes outputting the corrected point cloud data to the autonomous vehicle.