Vehicle Perception System Using Redundant Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated and semi-automated vehicle systems face reliability issues due to complex computerized control and uncertainty in operating environments, particularly in situations where global positioning systems are obstructed, leading to inaccurate localization and obstacle detection.
Innovation Solution
A system utilizing a versatile robotic control module with a distributed knowledge base and redundant sensors for high integrity perception, allowing vehicles to adapt movement based on dynamic conditions and sensor data fusion, enabling accurate navigation and obstacle avoidance even in environments with limited sensor visibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If global positioning systems are used for vehicle localization, then navigation capability is improved, but reliability deteriorates when GPS signals are obstructed
Solution Approach 1:
The system transitions from a centralized GPS-dependent localization approach to a distributed sensor network where each sensor (cameras, LIDAR, ultrasonic sensors) independently processes local environmental data to determine vehicle position and obstacles, ensuring reliable operation in GPS-denied environments
Solution Approach 2:
The localization system is segmented into multiple independent sensor modules, each capable of autonomous operation. The vehicle uses distributed sensors including forward-facing cameras, LIDAR, and ultrasonic sensors that can independently provide localization and obstacle detection without relying on a single GPS source
2Measurement precision
If multiple redundant sensors are deployed for high integrity perception, then measurement precision is improved, but device complexity increases
Solution Approach 1:
Multiple redundant sensors (cameras, LIDAR, ultrasonic sensors) are merged into a unified sensor processing system that combines their data streams. The system integrates visual data from cameras, range data from LIDAR, and proximity data from ultrasonic sensors to create a comprehensive environmental model, achieving high measurement precision while managing complexity through unified processing
Solution Approach 2:
The sensor system is designed with multi-functionality where the same sensor suite serves multiple purposes: localization, obstacle detection, and environmental mapping. The distributed sensors can independently perform different functions or work together, providing universal coverage for various operational scenarios without requiring separate specialized systems
3Productivity
If automated control systems are implemented, then productivity is improved, but reliability deteriorates due to system complexity
Solution Approach 1:
The control system implements dynamic switching between automated and manual operation modes. The system can automatically navigate using sensor data fusion for high productivity, but seamlessly transitions to manual control when reliability concerns arise or unexpected conditions occur, providing adaptive operational flexibility
Solution Approach 2:
The system incorporates continuous feedback loops where sensor data is constantly monitored and fed back to the control module. This real-time feedback enables the automated system to detect anomalies, verify its own decisions, and switch to manual mode when confidence thresholds are not met, maintaining reliability while enabling automated operation
Data Source
AI summary
The illustrative embodiments provide a computer program product for processing sensor data and controlling the movement of a vehicle. In an illustrative embodiment, an operating environment around the vehicle is identified and sensor data is selected from a set of sensors. A dynamic condition is identified using a plurality of different types of sensors on the vehicle. In response to the dynamic condition being identified, the movement of the vehicle is controlled. In another illustrative embodiment, an environment around the vehicle is identified to form an operating environment. Sensor data is selected from a set of sensors in a plurality of redundant sensors based on the operating environment.