Collaborative Occupancy Grid Fusion for Dynamic Environment Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing perception techniques based on occupancy grids are limited in accurately representing dynamic environments when multiple carriers with sensors are moving, leading to high uncertainties in occupancy state estimation, as each carrier computes a limited portion of the environment independently and frequently recomputes grids to maintain accuracy.
Innovation Solution
A collaborative method that fuses local occupancy grids from multiple carriers using Bayesian fusion to create a global occupancy grid, optimizing the use of available information and reducing uncertainty by computing occupancy probabilities of overlapping regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If each carrier independently computes a local occupancy grid covering a limited portion of the environment, then the computational complexity for each carrier is reduced, but the measurement precision and reliability of the overall environmental perception deteriorates due to high uncertainties in occupancy state estimation
Solution Approach 1:
The environment is divided into multiple local occupancy grids, each computed independently by a different carrier. This segmentation allows each carrier to handle a manageable portion of the environment with reduced computational complexity, while the overall coverage is maintained through multiple local grids.
Solution Approach 2:
Multiple local occupancy grids from different carriers are merged into a single global occupancy grid. This combining process integrates information from multiple sources, reducing uncertainty and improving the reliability and measurement precision of the overall environmental perception.
2Measurement precision
If each carrier frequently recomputes its local occupancy grid to maintain accuracy in dynamic environments, then the measurement precision is improved, but the productivity and loss of time worsen due to frequent recomputations
Solution Approach 1:
The system maintains continuous environmental perception by having multiple carriers simultaneously compute and update their local occupancy grids. This continuous parallel computation ensures that the global occupancy grid remains accurate in dynamic environments without requiring frequent individual recomputations by each carrier.
Solution Approach 2:
Each carrier pre-computes its local occupancy grid based on its current position and sensor data. These preliminary local grids are then integrated into the global grid, allowing the system to maintain accuracy without requiring frequent full recomputations by each individual carrier.
3Device complexity
If each carrier independently computes local occupancy grids without collaboration, then the device complexity is reduced, but the loss of information increases as each carrier only exploits measurements from its own sensors
Solution Approach 1:
The global occupancy grid serves multiple functions: it provides environmental perception for all carriers simultaneously, integrates data from multiple sensor sources, and reduces uncertainty through collaborative fusion. This multi-functional approach maximizes the use of available information without significantly increasing individual carrier complexity.
Solution Approach 2:
The global occupancy grid acts as an intermediary that collects, integrates, and fuses information from multiple local occupancy grids. This mediator structure allows information from all carriers to be combined effectively, reducing information loss while maintaining a relatively simple overall system architecture.
Data Source
AI summary
A method is provided for collaboratively detecting tangible bodies in an environment employing a plurality of distance sensors with which respective carriers are equipped, including: for each sensor, acquiring a series of measurements of distance of a closest tangible body along a line of sight; applying an inverse model of the sensor on a local occupancy grid; and constructing a consolidated local occupancy grid via Bayesian fusion of the occupancy probabilities thus determined; and, computing occupancy probabilities of the cells of a global occupancy grid via Bayesian fusion of the occupancy probabilities of corresponding cells of at least certain of said consolidated local occupancy grids. A system for implementing such a method is also provided.