Sensor Fusion via Environmental Model Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated driving systems face challenges in accurately detecting and tracking objects across overlapping sensor fields due to independent object detection by each sensor, leading to inconsistencies and reduced confidence levels when sensors diverge in their measurements.
Innovation Solution
A computing system that implements sensor event detection and fusion by spatially and temporally aligning raw measurement data from multiple sensors to generate a fused detection event, enhancing object detection confidence and enabling more accurate vehicle control through an environmental model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If each sensor performs object detection separately based on its own measurements, then each sensor can independently detect objects within its field of view, but inconsistencies and reduced confidence levels occur when sensors diverge in their measurements
Solution Approach 1:
The patent merges measurements from multiple sensors at the raw measurement data level before object detection. The sensor fusion system combines temporally-aligned measurements from multiple sensors with overlapping fields of view, allowing the system to detect objects using fused measurements rather than separate sensor detections. This approach resolves inconsistencies between sensors and improves detection confidence by utilizing complementary information from multiple sensors simultaneously.
Solution Approach 2:
The patent introduces an environmental model as an intermediary structure that serves as a common reference frame for fusing measurements from multiple sensors. The environmental model stores and aligns measurements in a unified spatial and temporal framework, enabling consistent object detection across sensors with overlapping fields of view. This intermediary model mediates between individual sensor measurements and final object detection decisions, resolving divergences through systematic fusion.
2Area of stationary object
If multiple sensors are used to detect objects, then coverage of the environment is improved, but the complexity of integrating and fusing sensor data increases
Solution Approach 1:
The patent segments the sensor fusion process into distinct functional modules: temporal alignment of measurements, spatial mapping to environmental model coordinates, and object detection from fused measurements. By dividing the complex fusion task into manageable segments, the system can process data from multiple sensors systematically without overwhelming computational complexity. Each sensor's measurements are processed independently through the same segmentation pipeline, making the system scalable.
Solution Approach 2:
The patent performs preliminary temporal alignment and spatial mapping of sensor measurements to the environmental model before object detection. By pre-aligning measurements in time and space within the environmental model framework, the system eliminates the need for complex real-time synchronization during object detection. This preliminary organization of sensor data reduces the computational burden of fusing measurements from multiple sensors with different update rates and coordinate systems.
3Reliability
If sensors have overlapping fields of view, then object detection confidence can be increased through integration, but the difficulty of detecting and measuring objects increases due to divergent sensor measurements
Solution Approach 1:
The patent transforms sensor measurements into a unified parameter space within the environmental model, using consistent coordinate systems and temporal references. By changing the parameters of individual sensor measurements (coordinates, timestamps, uncertainty values) into the standardized environmental model framework, the system can directly compare and fuse measurements from different sensors. This parameter transformation resolves measurement divergences and enables straightforward confidence calculation based on agreement across sensors.
Data Source
AI summary
This application discloses a computing system to implement sensor event detection and fusion system in an assisted or automated driving system of a vehicle. The computing system can monitor an environmental model to identify spatial locations in the environmental model populated with temporally-aligned measurement data. The computing system can analyze, on a per-sensor basis, the temporally-aligned measurement data at the spatial locations in the environmental model to detect one or more sensor measurement events. The computing system can utilize the sensor measurement events to identify at least one detection event indicative of an object proximate to the vehicle. The computing system can combine the detection event with at least one of another detection event, a sensor measurement event, or other measurement data to generate a fused detection event. A control system for the vehicle can control operation of the vehicle based, at least in part, on the detection event.