Situational-Awareness View Using Multi-Source Object Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Riders in autonomous vehicles experience anxiety due to a lack of understanding of how the vehicle perceives its environment and difficulty in interfacing with it, exacerbated by incomplete or inaccurate sensor data representations.
Innovation Solution
Supplementing autonomous-vehicle sensor data with secondary data such as user-generated, map, and inertial data to create a comprehensive and accurate situational-awareness view, using machine learning for classification and crowd-sourced feedback to enhance object recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If autonomous vehicle sensor data is used to represent the environment, then the vehicle can perceive objects and obstacles, but the representation may be incomplete or inaccurate causing rider anxiety
Solution Approach 1:
The patent combines multiple data sources including sensor data from the autonomous vehicle, sensor data from the rider's mobile device, and map data to create a comprehensive situational awareness view. This merging of multiple information sources compensates for incomplete or inaccurate sensor data from either source alone, providing a more complete and accurate representation of the environment.
Solution Approach 2:
The mobile device acts as an intermediary between the autonomous vehicle's sensor system and the rider. It receives sensor data from the vehicle, supplements it with additional sensor data from the rider's own device, and processes this combined information to generate an accurate situational awareness display, mediating the information flow to reduce rider anxiety.
2Ease of operation
If a human driver is present, then riders can interface with and control the vehicle, but the vehicle cannot operate autonomously
Solution Approach 1:
The autonomous vehicle system provides self-service by automatically processing sensor data, generating situational awareness representations, and making driving decisions without human intervention. The system serves itself by having the mobile device and vehicle processors automatically analyze data and control vehicle operations, eliminating the need for a human driver while maintaining operational capability.
Solution Approach 2:
The system implements feedback loops where sensor data from both the vehicle and rider's mobile device continuously informs the situational awareness model, which in turn guides vehicle control decisions. This closed-loop feedback system enables autonomous operation by allowing the vehicle to self-correct and adapt based on environmental perception and rider input.
3Speed
If sensor data is processed in real-time, then the situational awareness view is updated continuously, but processing complexity and computational requirements increase
Solution Approach 1:
The processing system is segmented into multiple components: the autonomous vehicle's processors handle vehicle sensor data and basic environmental perception, while the rider's mobile device handles additional sensor processing and final situational awareness model generation. This segmentation distributes computational complexity across multiple devices, enabling real-time processing without overloading a single system.
Data Source
AI summary
In one embodiment, a method includes receiving sensor data corresponding to an environment external of a vehicle. The sensor data include data points. The method includes determining one or more subsets of the data points. The method includes comparing the one or more subsets of the data points to one or more predetermined data patterns. Each of the one or more predetermined data patterns corresponds to an object classification. The method includes computing a confidence score for each subset of data points of the one or more subsets of the data points as corresponding to each of the one or more predetermined data patterns based on the comparison. The method includes generating a classification for an object in the environment external of the vehicle based on the confidence score.