Robot Path Planning With Probabilistic Occupancy Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for autonomous robots face challenges in accurately determining occupancy levels and planning paths in dynamic environments due to inconsistencies and uncertainties in incremental mapping, which can lead to inaccurate maps and potential collisions.
Innovation Solution
The system determines the occupancy level and probability distribution function (PDF) of each voxel using incremental data, calculates a cost function based on the PDF, and simultaneously plans the path and maps the environment to minimize collision risks and optimize resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If an incremental approach is used for generating a map, then the robot can continuously update the map with new sensor measurements, but there is uncertainty in the occupancy levels due to independent sensor measurements and assumptions
Solution Approach 1:
The system uses a probabilistic framework where occupancy levels are continuously refined through Bayesian updates. Each new sensor measurement provides feedback that adjusts the probability distribution of occupancy levels, allowing the system to maintain reliable estimates despite the incremental nature of data collection. The feedback mechanism incorporates measurement likelihoods and prior beliefs to continuously improve occupancy level accuracy.
Solution Approach 2:
The patent transforms the deterministic occupancy level parameter into a probabilistic parameter (probability distribution function). This parameter change allows the system to represent uncertainty explicitly and update occupancy levels continuously as new measurements arrive, resolving the contradiction between continuous updating and accuracy by embracing probabilistic reasoning.
2Ease of operation
If the robot plans a path using conventional methods, then it can determine a trajectory, but it may collide with obstacles due to inaccurate occupancy level information
Solution Approach 1:
The system changes the path planning parameter from deterministic occupancy levels to probability distribution functions of occupancy levels. This allows the planner to evaluate trajectories based on expected collision probabilities rather than binary occupied/free assumptions, improving collision avoidance while maintaining path planning capability.
Solution Approach 2:
The probability distribution function serves as an intermediary between the incremental mapping process and the path planning process. It translates uncertain sensor measurements into a format that the path planner can use to make informed decisions, bridging the gap between continuous map updating and reliable collision avoidance.
3Measurement precision
If the robot gathers all sensor data before calculating the map, then the map can be generated at once, but the robot cannot operate in real-time dynamic environments
Solution Approach 1:
The system performs preliminary actions by maintaining a probabilistic map representation that is continuously updated with new measurements. Rather than waiting to gather all data, the robot pre-computes probability distributions that can be quickly queried for path planning, enabling real-time operation while maintaining measurement precision through ongoing refinement of the probabilistic model.
Solution Approach 2:
The patent implements a dynamic mapping approach where the map is represented as a probabilistic structure that evolves continuously with new measurements. This dynamic representation allows the system to balance map accuracy with real-time responsiveness, as the probabilistic model can be updated incrementally without requiring complete data collection before operation.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
A method substantially simultaneously plans a path and maps an environment by a robot. The method determines a mean of an occupancy level for a location in a map. The method also includes determining a probability distribution function (PDF) of the occupancy level. The method further includes calculating a cost function based on the PDF. Finally, the method includes simultaneously planning the path and mapping the environment based on the cost function.