Landmark-Based Obstacle Detection Using Digital Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing obstacle detection systems in vehicles fail to reliably detect all obstacles, particularly those with non-reflective surfaces or those that may be obscured by other objects, leading to potential collisions due to incomplete or inaccurate data from detection units like lidar and ultrasonic sensors.
Innovation Solution
A method utilizing a digital map with stored locations of landmarks and obstacles, where the detection area is determined based on known obstacles, and landmarks are used to confirm the presence or absence of obstacles within this range, ensuring comprehensive detection and preventing measurement errors by using characterizing properties and multiple landmarks arranged at specific distances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If detection units like lidar and ultrasonic sensors are used for obstacle detection, then detection capability is improved, but obstacles with non-reflective surfaces or those obscured by other objects cannot be detected reliably
Solution Approach 1:
The patent introduces landmarks as intermediary objects placed in the environment to mediate the detection process. Instead of relying solely on detection units to directly sense obstacles (which fail for non-reflective or obscured objects), the system uses landmarks as reference points that are reliably detectable. The presence or absence of landmarks in expected positions indirectly indicates obstacle locations, solving the problem of incomplete direct detection data.
Solution Approach 2:
The system establishes a feedback loop by comparing the actual detected landmark positions with the expected positions stored in the digital map. When landmarks are not detected at their expected locations or are detected at deviating positions, the system infers the presence of obstacles. This feedback mechanism compensates for the limitations of direct obstacle detection by using indirect evidence from landmark observations.
2Measurement precision
If landmarks are used to detect obstacles by checking their presence or position, then detection accuracy for non-reflective obstacles is improved, but the system complexity increases due to digital maps and multiple detection criteria
Solution Approach 1:
The patent applies preliminary action by pre-storing landmark positions and characteristics in a digital map before the detection process begins. This preparation work is done in advance, allowing the detection system to simply compare actual observations against the pre-established reference data. This reduces the computational complexity during real-time operation while maintaining high measurement precision through the use of predetermined landmark positions and characterizing properties.
Solution Approach 2:
The system changes the detection parameter from directly measuring obstacle properties (which is difficult for non-reflective objects) to measuring landmark positions and characteristics. By transforming the detection task into observing changes in landmark parameters (position, presence, detectability), the system achieves high precision obstacle location accuracy while using a relatively simple detection mechanism that leverages the already-stored digital map data.
3Area of stationary object
If multiple landmarks are arranged at specific distances to ensure detection coverage, then detection coverage is improved, but the cost and complexity of deploying and maintaining landmarks increases
Solution Approach 1:
The patent makes landmarks multi-functional by using them for both vehicle orientation/navigation and obstacle detection. The same landmark infrastructure that helps vehicles navigate and locate themselves in the environment also serves as a reference system for detecting obstacles. This universal use of landmarks means that the same objects serve multiple purposes, reducing the need for separate detection infrastructure and lowering the overall quantity of landmarks required compared to dedicated obstacle detection systems.
Data Source
Figure 1~2
Figure 3
AI summary
The method and computation unit (10) for recognising obstacles first scan a digital map with a stored location of at least one landmark, a current location of at least one detection unit (40), and additional locations of known instantaneous and/or permanently stationary obstacles. In the next method step, a possible detection area of the at least one detection unit (40) is determined. Then, the environment within the possible detection area of the at least one detection unit (40) is scanned depending on the stored location, within the detection area, of the at least one landmark. An obstacle is subsequently recognised if the landmark is not detected in the scanned environment within the possible detection area or at a location deviating from the location stored on the digital map.