Geo-Spatial Object Representation with Iterative Quality Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for monitoring widely distributed and loosely connected objects such as railway grids, electricity grids, or oil and gas pipelines are inefficient and error-prone, leading to issues like power outages, wildfires, derailments, and leaks due to inadequate detection of necessary measures, and involve high costs and long reaction times.

Innovation Solution

A fully automated method for processing geo-spatial data to derive spatial object representations, using artificial intelligence techniques like neural networks, and ensuring quality conditions are met through iterative data collection and processing, allowing for continuous improvement and reduced errors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual inspection methods are used for monitoring distributed objects, then operational control can be maintained, but monitoring efficiency is low and costs are high

Engineering Contradiction:
Improvemonitoring efficiencyVSAvoidreaction time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical inspection methods with automated optical measurement systems (satellite imagery, aerial photography, ground-based sensors) to monitor railway infrastructure. This substitution enables continuous automated detection of vegetation encroachment, track geometry changes, and other spatial objects, dramatically improving monitoring efficiency while reducing reaction times from days/weeks to hours or minutes.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service monitoring where the infrastructure itself is continuously observed by automated detection systems. The spatial object representations automatically update and alert operators to issues without requiring manual intervention, allowing the system to monitor itself and trigger maintenance workflows autonomously.

Inventive Principle:
Principle #25Self-service

2Measurement precision

If manual inspection processes are used, then operational control is possible, but detection accuracy is insufficient leading to errors

Engineering Contradiction:
Improvedetection accuracyVSAvoiderror rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

Manual visual inspection is replaced with automated image processing and neural network-based object recognition systems. These systems consistently identify spatial objects (vegetation, tracks, signals, bridges) with high accuracy and uniformity, eliminating human errors such as missed detections, misidentifications, and fatigue-related mistakes that plague manual inspection.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system implements feedback loops where detected spatial objects and their characteristics are continuously analyzed, and results are fed back into the monitoring system. Quality conditions are evaluated and fed back to trigger additional data collection or re-processing when detection confidence is insufficient, ensuring high accuracy through iterative refinement.

Inventive Principle:
Principle #23Feedback

3Area of stationary object

If comprehensive monitoring is implemented, then coverage is improved, but data quality consistency becomes difficult to maintain

Engineering Contradiction:
Improvemonitoring coverageVSAvoiddata quality consistency
Core Design Contradiction:
Area of stationary objectVSManufacturing precision

Solution Approach 1:

The system dynamically adjusts processing parameters based on the specific characteristics of each spatial object and data source. Different neural network models and processing algorithms are applied depending on the object type (vegetation, track, signal), sensor modality (optical, radar, LiDAR), and environmental conditions, ensuring consistent quality across diverse monitoring areas while maintaining adaptability to local variations.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

Quality conditions are systematically evaluated for all derived spatial object representations, and feedback mechanisms trigger re-processing or additional data collection when quality thresholds are not met. This ensures that despite the large and varied monitoring coverage, all data maintain consistent quality standards through automated quality control loops.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4088218B1Method and system for analyzing geo-spatial data to derive spatial object representations
Publication Date: 2025.08.27 LIVEEO GMBH
  • EP4088218B1 patent drawingFigure 1
  • EP4088218B1 patent drawingFigure 2
  • EP4088218B1 patent drawingFigure 3A~3B

AI summary

Embodiments related to a method for analyzing geo-spatial data (11) comprising: c. Processing (S1) at least parts of the geo-spatial data (11) to derive spatial object representations (12); d. Collecting (S3) further geo-spatial data (11') if a quality condition of the spatial object representation (11') is not fulfilled. The proposed embodiments further comprise a system and a computer program.