AI Module for Object Tracking in Automotive Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of creating appropriate sensor models for object tracking in the automotive field, particularly for broad objects like vehicles, hinders effective association and updating of object status using environment sensor signals.
Innovation Solution
A method utilizing an AI module trained to associate detected sensor values with objects, updating the object status within a Bayesian filter, specifically using a grid-based artificial neural network or Point T-Net to map sensor measurements into an abstract dimension for improved robustness and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sensor models are used for object tracking, then the system structure is simple, but the association accuracy and robustness are insufficient
Solution Approach 1:
The patent replaces traditional mechanical/mathematical sensor models with an AI-based neural network model. The neural network learns complex relationships between sensor measurements and object states through training, substituting conventional analytical models with a data-driven approach that achieves superior association accuracy and robustness for tracking broad objects like vehicles.
Solution Approach 2:
The patent transforms the sensor model from a fixed mathematical formulation to a neural network with learnable parameters. Through training on labeled data, the network adapts its internal parameters to optimize the mapping between sensor measurements and object states, enabling it to handle complex scattering patterns and improve association accuracy dynamically.
2Reliability
If complex sensor models are used to improve association accuracy, then the tracking robustness improves, but the computational complexity increases
Solution Approach 1:
The neural network model is pre-trained offline using a large dataset of labeled sensor measurements and object states. This preliminary training phase allows the network to learn optimal association patterns beforehand, so that during online tracking, the network only needs to perform forward inference rather than complex real-time learning, reducing computational burden while maintaining high robustness.
Solution Approach 2:
The patent creates a virtual copy of the physical sensor measurement process through the neural network. The network learns to replicate the complex relationship between raw sensor data and object states, providing a computational model that can be applied repeatedly without the physical complexity of traditional sensor systems.
3Measurement precision
If AI module is introduced to update object status, then the accuracy of object status updates improves, but the system complexity increases
Solution Approach 1:
The neural network serves as an intermediary component between the sensor measurements and the object status representation. It processes raw sensor data and transforms it into improved object status estimates, mediating the information flow and enabling accurate status updates without requiring direct complex interactions between all system components.
Solution Approach 2:
The neural network model is designed to handle multiple functions: it performs association of sensor measurements with objects, updates object status, and can adapt to different object types and scenarios. This multi-functionality reduces the need for separate specialized modules, managing system complexity while improving overall accuracy.
Data Source
AI summary
A method is for tracking an object using an environment sensor. The object is represented by an object status. The method includes detecting a sensor value of the environment sensor, predicting a future object status of the object, and updating the object status using a Bayesian filter. The updating includes using an artificial intelligence module (“AI module”). The AI module is trained such that the detected sensor value is associated with the object and the object status of the object is updated based on the predicted future object status of the object and the detected sensor value.