Surfel Map Path Planning for Dynamic Occlusion Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating dynamic environments due to occlusions caused by changing objects like vegetation, which traditional mapping techniques struggle to track, leading to difficulties in making accurate navigation and collision avoidance decisions.
Innovation Solution
The use of surfel maps combined with real-time sensor data allows vehicles to predict and adjust their path to account for occlusions by assigning uncertainty to both surfel and sensor data, leveraging a 3D representation of the environment to improve accuracy and flexibility in navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional mapping techniques are used to represent the environment, then the map structure is simple and easy to process, but the system cannot accurately track changing objects like vegetation over time, leading to occlusion problems
Solution Approach 1:
The patent transitions from traditional 2D map representations to a 3D surfel (surface element) map structure. Each surfel contains position, orientation, and semantic information, enabling the system to represent complex environmental features including vegetation and occluding objects in three-dimensional space. This dimensional enhancement allows accurate tracking of changing objects while maintaining computational tractability through structured data organization.
2Loss of information
If the vehicle relies solely on real-time sensor data for navigation, then the system is responsive to current conditions, but it cannot predict occlusions from objects that are currently out of sensor view
Solution Approach 1:
The system performs preliminary path planning using the surfel map to identify potential occlusions and dangerous regions before the vehicle reaches them. By pre-processing the 3D surfel data to detect occluding objects and predict their impact on sensor visibility, the system prepares alternative paths in advance, reducing real-time computational burden and enabling proactive collision avoidance.
Solution Approach 2:
The patent incorporates safety margins and uncertainty buffers into the path planning process. When potential occlusions are detected in the surfel map, the system expands the vehicle's predicted occupancy volume and adds temporal buffers to account for undetected objects. This beforehand cushioning compensates for limited sensor visibility and prevents collisions with occluded objects.
3Measurement precision
If the surfel map includes detailed 3D information about all objects, then occlusion detection accuracy improves, but computational load and data processing requirements increase
Solution Approach 1:
The system applies local quality by focusing computational resources on surfels relevant to the vehicle's current path and potential occlusions. Rather than processing the entire surfel map uniformly, the algorithm identifies and analyzes only those surfels within the vehicle's field of view and adjacent regions where occlusions might occur. This localized processing maintains high detection precision while reducing overall computational load.
Data Source
AI summary
Methods, systems, and apparatus for generation and use of surfel maps to plan for occlusions. One of the methods includes receiving a previously-generated surfel map depicting an area in which a vehicle is located, the surfel map comprising a plurality of surfels, each surfel corresponding to a respective different location in the area in which a vehicle is located; receiving, from one or more sensors, sensor data representing the area in which the vehicle is located; determining, based on the sensor data, that the area in which a vehicle is located includes a dynamic object having a changed shape relative to its representation in the surfel map; and generating an updated path for the vehicle to travel that avoids an occlusion by the changed shape of the dynamic object of a line of sight of one or more sensors to an area of interest.