Remote Vehicle Control via Sensor Fusion in Close-Quarter Environments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Remote operation of vehicles in close-quarter environments, such as parking lots, faces challenges due to the need for precise control and the limitations of relying solely on visual data or pre-mapped high-definition maps, which do not account for dynamic obstacles and changing conditions.
Innovation Solution
A computer system in the vehicle collects data from multiple sensors to generate a rich visual representation of the environment, including camera images, LIDAR, radar, and ultrasonic data, which is then transmitted to a remote operator for real-time control, enabling precise navigation and parking through sensor fusion and automated control systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If visual data from onboard cameras is used for remote operation, then the remote operator can see the environment, but the precision for determining distances between objects is insufficient
Solution Approach 1:
The patent combines multiple sensor types (cameras, LIDAR, radar, ultrasonic sensors) into an integrated sensor system that collects data from different modalities simultaneously. This merging of sensor data sources enables the remote operator to access both visual information and precise distance measurements, resolving the contradiction between information completeness and measurement precision.
Solution Approach 2:
The patent introduces a remote computer system as an intermediary between the vehicle's sensor system and the operator. This intermediary processes and fuses data from multiple sensors, then presents integrated environmental information to the remote operator, enabling precise distance determination while maintaining comprehensive environmental awareness.
2Measurement precision
If HD maps with accurate location are used, then distance determination precision is improved, but remote operation is limited to pre-mapped environments only
Solution Approach 1:
The patent transitions from static HD map-based positioning to a dynamic sensor fusion system that continuously adapts to changing environments. The system processes real-time data from multiple sensors to dynamically determine vehicle location and environmental features, enabling operation in both pre-mapped and unmapped environments while maintaining high precision.
Solution Approach 2:
The patent creates a universal remote operation system that can function across diverse environments by incorporating multiple sensor types capable of operating in various conditions. The sensor fusion approach allows the system to adapt to different environment types (parking lots, streets, garages) without requiring pre-mapping, while still achieving HD map-level precision through real-time data processing.
3Measurement precision
If sensors are integrated into the environment in advance, then distance measurement precision is improved, but the setup complexity and cost increase
Solution Approach 1:
The patent implements a self-service approach where the vehicle carries its own sensor system rather than relying on externally deployed environmental sensors. The vehicle's onboard sensors autonomously collect environmental data and determine distances without requiring pre-installation of sensors in the operating environment, reducing deployment complexity while maintaining high measurement precision.
Solution Approach 2:
The patent segments the sensing function across multiple independent sensor units (cameras, LIDAR, radar, ultrasonic sensors) distributed throughout the vehicle. This segmentation allows each sensor type to specialize in specific measurement tasks while the system as a whole achieves comprehensive environmental awareness with high precision, avoiding the need for a single complex sensor system or environmental infrastructure.
4Loss of information
If multiple sensors are used for sensor fusion, then environmental information completeness is improved, but the data processing complexity increases
Solution Approach 1:
The patent employs a remote computer system as an intermediary that handles the complex task of fusing data from multiple sensors. This intermediary processes raw data from cameras, LIDAR, radar, and ultrasonic sensors, integrating them into a coherent environmental model that is then presented to the remote operator, managing processing complexity while maximizing information completeness.
Solution Approach 2:
The patent extracts and separates the complex data fusion processing from the vehicle's onboard systems and relocates it to a remote computer system. This extraction allows the vehicle to collect comprehensive sensor data with minimal onboard processing burden, while the remote system handles the computationally intensive fusion operations, reducing the processing complexity at the vehicle level while maintaining complete environmental information.
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
This approach allows for centimeter-level accuracy in vehicle control, adapting to dynamic environments and unexpected obstacles, enhancing safety and efficiency in close-quarter operations by providing a robust and adaptive remote operation system.
Implementation Method 1
Light Detection and Ranging (LIDAR) sensor
Implementation Method 2
Radio Detection and Ranging (radar) sensor
Data Source
AI summary
Techniques for remote operation of vehicles are presented. Remote operation can involve maneuvering, by a vehicle computer system, a vehicle based on driving instructions from a remote computer system. The vehicle computer system collects sensor data from a plurality of sensors onboard the vehicle. Based on the sensor data, data associated with a visual representation of the surrounding environment is generated and sent to the remote computer system. The generating of the data can involve fusing sensor data to identify one or more features of a surrounding environment. The generated data may be indicative of the identified features. The visual representation is output on a display device of the remote computer system. The driving instructions can be based on human input supplied in response to the visual representation. In certain embodiments, the vehicle is maneuvered toward a location within proximity of a parking space before performing automated parking.