Collaborative UGV-UAV Surveillance for Real-Time Anomaly Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current surveillance systems for facilities lack effective anomaly detection and classification, particularly in dynamic environments, as they struggle to accurately differentiate between critical and non-critical states, often resulting in false alarms and requiring extensive manual intervention.
Innovation Solution
A facility surveillance system utilizing a dynamic building information model (BIM) with integrated sensors and a central computing unit that employs machine learning and multi-modal sensor fusion to classify state patterns based on topological, logical, and functional relationships, enabling real-time anomaly detection and classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional surveillance systems are used for facility monitoring, then the system structure is simple, but the anomaly detection accuracy is low and false alarms occur frequently
Solution Approach 1:
The patent combines multiple sensor types (motion detectors, temperature sensors, humidity sensors, light sensors, cameras) into an integrated surveillance system that collects and processes data from diverse sources simultaneously, enabling more accurate anomaly detection through multi-parameter analysis rather than relying on single-sensor data
Solution Approach 2:
The system dynamically adjusts surveillance parameters based on learned patterns from historical data and real-time conditions, adapting sensitivity thresholds and detection criteria to reduce false alarms while maintaining high detection accuracy for actual anomalies
2Reliability
If traditional surveillance systems generate alarms for all detected anomalies, then all potential issues are flagged, but the number of false alarms increases and manual intervention is required
Solution Approach 1:
The system implements feedback loops where alarm history, false alarm patterns, and manual verification results are fed back into the machine learning models, continuously refining the classification algorithms to better distinguish between critical anomalies and false alarms, thereby improving alarm accuracy over time
Solution Approach 2:
The surveillance system performs self-classification of detected anomalies using integrated machine learning models that automatically evaluate sensor data patterns, correlate multiple sensor inputs, and determine whether detected states represent true anomalies or false alarms, reducing the need for manual verification
3Area of stationary object
If comprehensive surveillance of all facility elements is implemented, then complete monitoring coverage is achieved, but the data processing load and system complexity increase
Solution Approach 1:
The facility is divided into discrete spatial zones and logical areas, with surveillance data processed and classified at local levels before being aggregated to the central system, enabling comprehensive monitoring coverage while distributing computational load and reducing central processing complexity
Data Source
Figure 1~2a
Figure 2b~2c
Figure 2d~3
AI summary
The invention pertains to a mobile surveillance system (200) adapted for patrolling a surveillance area of a facility like a building, the system comprising a plurality of sensors and at least one unmanned ground vehicle (UGV; 210) that is adapted to move autonomously on a ground of the surveillance area, the UGV (210) comprising a housing (211) enclosing - a first battery (217); - first sensor means, particularly comprising a first camera (213), the first sensor means being adapted to generate first sensor data, - a first computing unit (218) comprising a processor and a data storage, the first computing unit (218) being adapted to receive and evaluate the first sensor data in real time. According to the invention, the system (200) further comprises at least one unmanned aerial vehicle (UAV; 220), the UGV (210) and the UAV (220) being adapted for collaboratively patrolling the surveillance area, in particular autonomously, wherein - the UAV (220) comprises second sensor means (223), particularly comprising a second camera (223), the second sensor means being adapted to generate second sensor data (252), - the UGV (210) comprises a first data exchange module (215) and the UAV (220) comprises a second data exchange module (225), the first and second data exchange modules (215, 225) being adapted to exchange data (251, 252); and - the first computing unit (218) is adapted to receive and evaluate the second sensor data (252) in real time.