Navigation Data Generation via Experience Node Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems face challenges in maintaining long-term autonomy in changing environments, such as sudden structural changes, lighting conditions, and seasonal changes, which can lead to limited success in re-localization and unnecessary experience creation due to the lack of effective linking between experiences in a plastic mapping system.
Innovation Solution
A method is introduced to generate navigation data by receiving new and stored experience data sets, identifying candidate nodes for matching, and creating topological links between experiences, which are then validated using robust matching techniques to enhance the connectivity of experiences and reduce unnecessary experience creation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a visual navigation system uses a single frame of reference for mapping, then localization accuracy is improved, but the system fails to maintain long-term autonomy in changing environments
Solution Approach 1:
The system dynamically switches between multiple frames of reference (local maps) based on environmental conditions and vehicle location. Each local map maintains its own frame of reference, allowing the system to adapt to environmental changes while preserving localization accuracy within each context. The plurality of local maps can be actively switched or blended depending on which best matches current conditions.
Solution Approach 2:
The global mapping space is segmented into multiple local maps, each covering a specific temporal or spatial segment of the environment. This segmentation allows each local map to maintain independence with its own frame of reference, reducing the impact of environmental changes on overall localization. The system divides the navigation problem into manageable local segments rather than requiring a single comprehensive global map.
2Adaptability or versatility
If the system creates new experiences frequently in response to environmental changes, then adaptability is improved, but computational resources are wasted through unnecessary experience creation
Solution Approach 1:
The system performs preliminary comparison of new sensor data against existing local maps before committing to create new experiences. By pre-evaluating whether environmental changes exceed predefined thresholds, the system avoids unnecessary experience creation and computational overhead. Only when changes are significant enough to warrant a new local map does the system proceed with experience creation.
Solution Approach 2:
The system continuously monitors environmental changes and provides feedback to the experience management logic. This feedback mechanism allows the system to distinguish between transient variations and significant environmental changes, triggering experience creation only when necessary. The feedback loop prevents premature or unnecessary experience generation while maintaining responsiveness to genuine environmental transformations.
3Reliability
If the system stores multiple experiences covering the same physical space, then re-localization capability is improved, but data redundancy increases
Solution Approach 1:
Each local map is optimized with specific quality characteristics suited to its temporal or spatial context. Rather than storing identical comprehensive data across multiple experiences, each local map contains selectively relevant information tailored to its specific context. This local optimization reduces redundancy while maintaining re-localization capability through contextual appropriateness.
Solution Approach 2:
The system selectively discards redundant information when creating new local maps that cover overlapping physical spaces. By identifying and eliminating duplicate or superseded data, the system maintains re-localization capability through essential unique information while reducing overall data redundancy. Old experiences are not completely discarded but their useful components are recovered and integrated into updated local maps.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Navigation data is generated by receiving (502) a new experience data set (321 ) relating to a new experience capture. At least one stored experience data set (320) relating to at least one previous experience capture is also received (504). An experience data set includes a set of nodes, with each node comprising a series of visual image frames taken over a series of time frames. A candidate set of said nodes is obtained (506) from the stored experience data set that potentially matches a said node in the new experience data set, and then a check (508) if performed to see if a node in the candidate set matches the node in the new experience data set. If the result of the checking is positive then data relating to the matched nodes is added (510) to a place data set useable for navigation, the place data set indicating that said nodes in different said experience data sets relate to a same place.