Occupancy Grid Mapping With Asynchronous Multi-Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional occupancy grid maps in AI applications, such as self-driving vehicles, are inaccurately updated due to limitations in laser radar sensors, leading to nondeterministic and unreliable maps, especially when these sensors malfunction.
Innovation Solution
A method and apparatus for creating an occupancy grid map using multiple sensors with an asynchronous updating policy, combining environment perception information from various sensors to improve accuracy and timeliness, including vehicle-mounted cameras and laser radars, to enhance the stability and real-time features of the map.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If laser radar is used as the sole sensor for updating occupancy grid map, then the map updating process is simple, but the map accuracy and reliability deteriorate due to short range and sparse detection results
Solution Approach 1:
The patent combines multiple sensors (laser radar, cameras, ultrasonic sensors, millimeter-wave radar) into a unified occupancy grid map updating system. Each sensor type contributes its strengths: laser radar for range and precision, cameras for visual recognition, ultrasonic sensors for short-range detection, and millimeter-wave radar for all-weather capability. This merging resolves the contradiction by achieving high map accuracy through multi-sensor fusion while managing system complexity through modular architecture and asynchronous updating policies.
Solution Approach 2:
The occupancy grid map system is designed to accept and process data from multiple types of sensors simultaneously, making the system multi-functional. The same occupancy grid map structure and updating algorithms can accommodate different sensor types with varying detection ranges, resolutions, and update frequencies. This universality allows the system to leverage diverse sensor capabilities without requiring separate processing pipelines for each sensor type.
2Reliability
If laser radar is used as the sole sensor, then the system complexity is low, but the system reliability deteriorates when the sensor malfunctions
Solution Approach 1:
The patent implements redundancy by deploying multiple sensor types that can compensate for each other's failures. If one sensor malfunctions, the system continues to operate using data from remaining functional sensors. The asynchronous updating policy allows the system to gracefully handle sensor failures by adjusting update frequencies and weights of available sensors, ensuring continuous occupancy grid map updates even when individual sensors fail.
Solution Approach 2:
The system dynamically adjusts the weighting and update frequency parameters of different sensors based on their operational status and environmental conditions. When a sensor malfunctions or performs poorly, its weight in the fusion algorithm is reduced or set to zero, and other sensors compensate by increasing their contribution. This parameter adaptation maintains system reliability without requiring complex failure detection and isolation mechanisms.
3Loss of time
If synchronous updating policy is used with multiple sensors, then the system structure is simple, but the map timeliness deteriorates due to different sensor update frequencies
Solution Approach 1:
The patent implements an asynchronous updating policy where each sensor type updates the occupancy grid map at its optimal frequency based on its detection capabilities and the dynamics of the environment it monitors. Fast-response sensors like ultrasonic sensors update frequently for near-range obstacles, while slower sensors like millimeter-wave radar update at lower frequencies for far-range detection. The system dynamically adjusts update frequencies and fuses results in real-time, achieving high map timeliness without requiring all sensors to synchronize their update cycles.
4Productivity
If multiple sensors with different update frequencies are used, then the map timeliness improves, but the system complexity increases
Solution Approach 1:
The patent segments the occupancy grid map updating process into independent sensor processing modules, each handling a specific sensor type with its own update frequency and processing logic. This segmentation allows each sensor to operate independently at its optimal update rate without coordinating with others, improving overall updating efficiency. The modular architecture manages system complexity by isolating processing logic into separate, manageable units that can be independently configured and maintained.
Data Source
AI summary
The present disclosure provides a method and an apparatus for creating an occupancy grid map, as well as a processing apparatus. The method includes: creating a current occupancy grid map based on a location of the vehicle and a previous occupancy grid map; and determining a current probability that each grid in the current occupancy grid map belongs to each of occupancy categories based on last environment perception information received from the sensors and updating an occupancy category to which each grid in the current occupancy grid map belongs based on the current probability that the grid belongs to each of the occupancy categories, in accordance with an asynchronous updating policy.