Bulldozer Camera SLAM for Autonomous Terrain Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current geography-altering machines, such as bulldozers, lack efficient and accurate methods for generating terrain information and navigating terrain sections autonomously, relying heavily on operator intervention and satellite positioning systems.
Innovation Solution
A control system equipped with at least one camera that captures image series while moving, utilizing SLAM or stereo photogrammetry algorithms to derive terrain data, including point clouds and position information, enabling autonomous navigation and precise terrain shaping without operator intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If satellite positioning systems are used to determine terrain information, then position information can be obtained, but the precision and detail of terrain data are insufficient for accurate terrain shaping
Solution Approach 1:
The patent replaces satellite positioning systems with a camera-based visual SLAM system. Instead of relying on satellite signals, the system uses cameras to capture images and algorithms to compute terrain information and position data, achieving higher precision without the limitations of satellite systems
Solution Approach 2:
The patent introduces cameras as intermediary devices between the bulldozer and the terrain. The cameras capture visual information that is then processed through SLAM algorithms to generate precise terrain models and position data, serving as a mediator that bridges the gap between the vehicle and the work environment
2Productivity
If operator intervention is required for terrain surveying and navigation, then flexibility is maintained, but productivity and efficiency are reduced
Solution Approach 1:
The patent implements autonomous operation where the bulldozer performs terrain surveying, positioning, and navigation automatically without operator intervention. The visual SLAM system and control algorithms enable the vehicle to service itself by independently capturing images, computing terrain information, and executing navigation commands
Solution Approach 2:
The patent establishes a feedback loop where the camera continuously captures terrain images, the SLAM algorithm processes this data to update position and terrain model information, and this feedback is used to automatically adjust the bulldozer's navigation and blade control, enabling closed-loop autonomous operation
3Loss of information
If conventional monitoring systems are used to observe material transport and work cycles, then basic functionality is provided, but real-time terrain information and navigation capabilities are lacking
Solution Approach 1:
The patent makes the camera system multi-functional by using it not only for observing work cycles and material transport but also for capturing terrain information, determining position, and generating navigation data. This single camera system serves multiple purposes, eliminating the need for separate monitoring and navigation systems
Solution Approach 2:
The patent merges the terrain surveying function with the navigation function. Both terrain information generation and position determination are achieved through the same visual SLAM processing of camera images, combining what were previously separate functions into a unified system
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system provides precise terrain information and navigation capabilities, allowing for efficient and accurate shaping of the terrain, enabling autonomous operation and real-time updates across a network of machines, enhancing construction efficiency and safety.
Implementation Method 1
at least one camera for capturing images of the environment, particularly of the terrain
Implementation Method 2
applying a simultaneous localisation and mapping (SLAM) algorithm to the set of image data and thereby deriving terrain data
Implementation Method 3
applying a stereo photogrammetry algorithm to the set of image data and thereby deriving terrain data
Data Source
Figure 1a~1d
Figure 2a~2d
Figure 3
AI summary
A Method for generating scaled terrain information while operating a bulldozer (1), the bulldozer (1) comprising a driving unit comprising a set of drive wheels (5), particularly being part of a continuous track (6), and a motor connected to at least one of the drive wheels (5) for providing movability of the bulldozer (1), a blade (4) for altering the surface of the terrain (10,10'), at least one camera (2,2',2",2"') for capturing images of the environment, particularly of the terrain (10,10'), the camera (2,2',2" ,2" ') being positioned and aligned in a known manner relative to the bulldozer (1), and a controlling and processing unit (3), wherein the method comprises: moving the bulldozer (1) in the terrain (10,10') whilst concurrently generating a set of image data by capturing an image series of terrain sections with the at least one camera so that at least two images, particularly successive images, of the image series cover an amount of identical points in the terrain (10,10'), wherein the terrain sections are defined by a viewing field (7,7',7",7"') of the at least one camera (2,2',2" ,2" ') at respective positions of the camera (2,2',2" ,2" ') while moving, and either applying a simultaneous localisation and mapping (SLAM) algorithm or a stereo photogrammetry algorithm to the set of image data and thereby deriving terrain data, the terrain data comprising a point cloud representing the captured terrain (10,10') and position data relating to a relative position of the bulldozer (1) in the terrain (10,10'), and scaling the point cloud by applying an absolute scale information to the terrain data, particularly scaling the position data.