Autonomous Vehicle Mapping for Dynamic Confined Industrial Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mapping techniques struggle to create accurate maps for autonomous vehicles in confined industrial areas where available driving areas change over time due to the presence of dynamic objects like moving containers or vehicles.
Innovation Solution
A system comprising a truck, server, and drone is used to obtain and classify environmental information, distinguishing between static, semi-static, and dynamic objects to create a dynamic map that reflects the current available driving area, enabling the truck to plan an optimal route.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional mapping techniques are used to create maps for autonomous vehicles, then map creation is simple and fast, but the map accuracy deteriorates in confined industrial areas where objects move frequently
Solution Approach 1:
The patent applies the Dynamics principle by creating a dynamic map that is continuously updated based on the movement of objects in the confined area. Instead of using a static map, the system dynamically adjusts the map to reflect current object positions, particularly for semi-static objects that move occasionally. This resolves the contradiction by making the mapping system adaptive to changing environments while maintaining reasonable complexity through automated detection and classification processes.
Solution Approach 2:
The patent implements feedback mechanisms by continuously monitoring object positions and using this information to update the map. The system detects object movements, classifies objects as static, semi-static, or dynamic, and feeds this information back into the mapping process. This feedback loop ensures map accuracy is maintained in environments with moving objects, resolving the contradiction between map accuracy and system complexity.
2Reliability
If multiple sets of attribute information are collected to improve map reliability, then the reliability of static objects improves, but the system cannot adapt to environments where objects change location over time
Solution Approach 1:
The patent applies segmentation by dividing objects into three distinct categories: static objects, semi-static objects, and dynamic objects. This segmentation allows the system to apply different reliability assessment and mapping strategies to each category. Static objects benefit from multiple attribute information collection for high reliability, while semi-static and dynamic objects are handled with location-tracking mechanisms, thus adapting to environments where objects change location.
Solution Approach 2:
The patent changes the parameter of object classification from a single static category to multiple dynamic categories (static, semi-static, dynamic). This parameter change enables the system to adapt to different object behaviors in the environment. By adjusting how objects are classified and tracked based on their movement characteristics, the system maintains reliability for stationary objects while adapting to environments where objects move, resolving the contradiction between reliability and environmental adaptability.
3Ease of manufacture
If a static map is created assuming objects remain at the same location, then map creation is straightforward, but the map becomes inaccurate when objects like containers or vehicles move
Solution Approach 1:
The patent transitions from a static map creation approach to a dynamic map updating approach. While the initial map creation remains straightforward, the system continuously updates the map to reflect object movements. This is achieved by detecting object positions, classifying them by mobility, and adjusting the driving area identification accordingly. This resolves the contradiction by maintaining ease of initial map creation while improving ongoing accuracy through automated dynamic updates.
Data Source
AI summary
When an autonomous vehicle plans a new drive mission in a confined area, a first set of environmental information, and after a time interval, a second set of environmental information, are obtained by a device flying above a confined area including the departure point and the destination point of the vehicle, each set of environmental information including locations and geometries of objects in the confined area. Then, an object that is present at the same location in the first set of environmental information and in the second set of environmental information is classified as a semi-static object, and an object that is present at different locations in the first set of environmental information and in the second set of environmental information is classified as a dynamic object. A map including location and geometry of each object classified as a semi-static object is then created.