Radar Point Cloud Seat Occupancy Detection for Stationary Passengers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing seat occupancy detection systems in vehicles face challenges in accurately distinguishing between different occupancy states and are limited by the need for pre-installed sensors, which hinder retrofitting and fail to detect occupants who are not moving significantly, leading to errors in seat occupancy status determination.
Innovation Solution
A method utilizing radar point clouds segmented into clusters based on spatial position, with an evaluation model analyzing radar points and Doppler shift values to determine seat occupancy states, allowing for reliable detection of seat occupancy status without the need for pre-installed sensors and reducing error rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If radar point clouds are used to detect seat occupancy, then the need for pre-installed sensors is eliminated and retrofitting becomes possible, but the accuracy of occupancy detection deteriorates when occupants are not moving significantly
Solution Approach 1:
The patent segments the radar point cloud into multiple clusters, each corresponding to a specific seat region. This segmentation allows the system to analyze occupancy in each seat independently, improving detection accuracy even for stationary occupants by comparing point distribution patterns within each segmented region rather than treating the entire detection area as a single unit.
Solution Approach 2:
The patent introduces Doppler shift values as an additional dimension for analyzing radar points. By evaluating both spatial position and Doppler shift characteristics, the system can distinguish between genuine occupancy and false detections, while also being able to detect stationary occupants through spatial pattern recognition alone when Doppler signals are absent or minimal.
2Reliability
If multiple radar point clouds from different measurement frames are used to improve detection reliability, then the accuracy of occupancy status determination is improved, but the processing time and computational load increase
Solution Approach 1:
The patent performs preliminary clustering of radar points into seat-specific regions before full occupancy analysis. By pre-organizing the point cloud data into spatial clusters corresponding to individual seats, the system reduces the computational complexity of subsequent multi-frame analysis, enabling faster processing while maintaining improved reliability through temporal comparison.
Solution Approach 2:
The patent segments both the spatial domain (into seat-specific clusters) and the temporal domain (into measurement frames). This dual segmentation allows the system to process multiple time frames efficiently by comparing clustered point distributions across time, improving reliability through temporal redundancy while reducing overall computational load through structured organization.
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
The method provides accurate and reliable detection of seat occupancy status, enabling improved vehicle functionality control, such as seat belt warnings and airbag activation, by segmenting radar point clouds and using an evaluation model to analyze radar points and Doppler shift values, thus overcoming the limitations of existing systems.
Implementation Method 1
each of which has one or more measurement frames of the same temporal length... obtained on the basis of a radar scan of a spatial area surrounding the seating arrangement
Implementation Method 2
The individual radar points of each radar point cloud are further represented by a Doppler shift value of the radar signal to the respective radar point
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A method for the automated detection of a seating arrangement's occupancy state comprises: receiving or generating measurement data, each of which represents an associated radar point cloud for each measurement frame from a set of multiple measurement frames, such that the measurement data represents a plurality of radar point clouds corresponding to the multiple measurement frames, the time duration of measurement frames from different groups of measurement frames differs, and each of the multiple groups of measurement frames is assigned a corresponding group of radar point clouds; and determining a seating arrangement's occupancy state using an evaluation model that, depending on at least two groups of radar point clouds, delivers one of several predefined possible seating arrangement occupancy states as an evaluation result;wherein for each radar point cloud the set of its radar points is segmented into clusters in order to individually assign each seat to the cluster closest to it spatially; for each cluster of at least one radar point cloud of the at least two groups of radar point clouds that is assigned to the measurement frame with the shortest time length, a number of radar points contained in the respective cluster or a quantity dependent thereon is determined, a value based on the number of radar points contained in the respective cluster or the quantity dependent thereon is compared with a first threshold value, a stored seat occupancy state is updated based on the evaluation result only if the value is less than the threshold value, and information defined as a function of the evaluation result is output when the stored seat occupancy state has been updated.