Occupancy Grid Initialization Using High-Density Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles rely on occupancy grid maps initialized to an unknown state, which can lead to inefficiencies in vehicle movement and navigation, as they gradually update cell states based on sensor data, lacking initial precision for static environmental features.
Innovation Solution
Initializing occupancy grid map cells with information from high density maps, which include static environmental features, to provide an estimated state, and updating based on sensed data to improve navigation and control vehicle movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If occupancy grid map cells are initialized to an unknown state, then the system maintains simplicity and uniformity in initialization, but the navigation accuracy and efficiency deteriorate due to lack of initial precision for static environmental features
Solution Approach 1:
The system performs preliminary action by initializing occupancy grid map cells with estimated occupancy information from high-density maps before the autonomous vehicle begins navigation. This pre-initialization provides an estimated state for static environmental features, improving navigation accuracy without requiring complex real-time processing during operation
Solution Approach 2:
The system introduces an intermediary mechanism by using high-density maps as a source of estimated occupancy information. These maps serve as a mediator between the initialization process and the occupancy grid map, providing preprocessed environmental data that improves initial precision without directly complicating the core navigation system
2Productivity
If occupancy grid map cells are updated gradually based on sensor data, then the system maintains simplicity in data processing, but the time required to confirm static features increases
Solution Approach 1:
The system performs preliminary action by pre-populating occupancy grid map cells with estimated occupancy information from high-density maps before navigation begins. This eliminates the need to gradually discover static features through sensor data collection, immediately reducing the time required to confirm static environmental features while maintaining simple sensor-based update processes
Solution Approach 2:
The system applies local quality by differentiating between static and dynamic features in the initialization process. Static features are initialized with estimated occupancy information from high-density maps for immediate confirmation, while dynamic features continue to be updated gradually through sensor data, optimizing navigation efficiency without overwhelming the processing system
3Measurement precision
If occupancy grid map cells are initialized with estimated occupancy information from high density maps, then the initial precision for static environmental features is improved, but the complexity of data processing and integration increases
Solution Approach 1:
The system extracts only the necessary estimated occupancy information from high-density maps for initializing occupancy grid map cells. Rather than processing and integrating all data from high-density maps, the system selectively extracts relevant occupancy information, reducing data processing complexity while maintaining initial precision for static environmental features
Solution Approach 2:
The system applies partial action by initializing only those occupancy grid map cells that correspond to static environmental features using estimated occupancy information from high-density maps. Dynamic features and areas beyond the vehicle's current sensor range are not initialized, reducing data processing complexity while still improving initial precision where it matters most for navigation
4Adaptability or versatility
If the system uses sensors to detect and transmit temporary obstacle information, then the ability to handle dynamic environmental changes is improved, but the loss of time for data transmission and processing increases
Solution Approach 1:
The system rushes through the data transmission process for temporary obstacles by immediately transmitting sensed occupancy information to the server when obstacles are detected, rather than waiting for complete data collection or processing cycles. This reduces data transmission time and improves adaptability to dynamic environmental changes
Solution Approach 2:
The system implements self-service by having each autonomous vehicle independently detect, process, and transmit temporary obstacle information to the server. This distributed approach eliminates centralized processing delays and enables rapid response to dynamic environmental changes, with each vehicle autonomously managing its own obstacle detection and communication
Data Source
AI summary
Systems and methods for controlling an autonomous vehicle. One system includes an electronic processor and a memory configured to store an occupancy grid map. The occupancy grid map includes a set of cells associated with respective positions in an environment. A sensor is configured to output sensed occupancy information associated with positions in the environment. The electronic processor is configured to determine a position of the autonomous vehicle within the environment, and receive, from a server, a high density map. The high density map contains occupancy information. The electronic processor initializes the cells with occupancy information based on the high density map. The electronic processor confirms, based on the sensed occupancy information, a state of the set of the cells of the occupancy grid map, and controls vehicle movement based on the occupancy grid map.