Phase-Coherent LiDAR Object Classification for Autonomous Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately collecting and interpreting environmental data to safely navigate complex environments, particularly in determining the classification, pose, and movement parameters of objects in their surroundings.
Innovation Solution
Utilizing phase coherent LIDAR data to determine the classification, pose, and instantaneous velocity of objects by processing LIDAR data points with a trained machine learning model, enabling adaptive autonomous control of the vehicle based on these determinations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phase coherent LIDAR data is used to determine object classification and movement parameters, then measurement precision and reliability are improved, but device complexity increases
Solution Approach 1:
The LIDAR data processing is segmented into distinct functional modules: data acquisition from phase coherent LIDAR, velocity calculation using Doppler shift analysis, classification model processing, and control signal generation. Each module handles a specific aspect of the overall task, making the complex system more manageable and maintainable while preserving measurement precision.
Solution Approach 2:
A trained classification model serves as an intermediary between the raw LIDAR data and the vehicle control system. This intermediary processes the complex phase coherent data, extracting relevant features and making classification decisions, thereby simplifying the interface between sensing and actuation while maintaining high measurement precision.
2Reliability
If phase coherent LIDAR data processing is used to determine instantaneous velocity and classification, then reliability of autonomous control is improved, but loss of time in processing increases
Solution Approach 1:
The classification model is trained in advance on large datasets of LIDAR data with known object classifications and movement parameters. This preliminary training phase allows the model to learn complex patterns and decision boundaries, enabling rapid and reliable real-time classification during actual autonomous operation without requiring extensive processing time during critical moments.
Solution Approach 2:
Traditional mechanical or rule-based velocity measurement and object classification methods are replaced with a machine learning-based classification model. This substitution enables parallel processing of multiple LIDAR data points and features simultaneously, significantly reducing processing time while improving reliability through pattern recognition capabilities that capture complex object behaviors.
3Measurement precision
If subgroup of LIDAR data points corresponding to objects is determined using both range and velocity, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system transitions from analyzing LIDAR data in a single dimension (range only) to utilizing multiple dimensions simultaneously (range, velocity, and derived classification features). By incorporating velocity information from phase coherent measurements and feeding both range and velocity data into the classification model, the system achieves more precise object detection and classification while the modular architecture manages the increased processing complexity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables precise and timely adaptation of vehicle control to avoid collisions with objects by accurately classifying and tracking their movement, enhancing safety and navigation in various driving conditions.
Implementation Method 1
receiving, from a phase coherent Light Detection and Ranging (LIDAR) component of a vehicle, a group of phase coherent LIDAR data points collectively capturing a plurality of points in an area of an environment of the vehicle. Each of the phase coherent LIDAR data points of the group indicate a corresponding range and a corresponding velocity
Implementation Method 2
Each of the phase coherent LIDAR data points of the group indicate a corresponding range and a corresponding velocity for a corresponding one of the points in the environment
Data Source
AI summary
Determining classification(s) for object(s) in an environment of autonomous vehicle, and controlling the vehicle based on the determined classification(s). For example, autonomous steering, acceleration, and/or deceleration of the vehicle can be controlled based on determined pose(s) and/or classification(s) for objects in the environment. The control can be based on the pose(s) and/or classification(s) directly, and/or based on movement parameter(s), for the object(s), determined based on the pose(s) and/or classification(s). In many implementations, pose(s) and/or classification(s) of environmental object(s) are determined based on data from a phase coherent Light Detection and Ranging (LIDAR) component of the vehicle, such as a phase coherent LIDAR monopulse component and/or a frequency-modulated continuous wave (FMCW) LIDAR component.