UAS Threat Assessment Using Multi-Source Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The detection and threat assessment of small unmanned aerial systems (UASs) operating at low altitudes or within limited lines of sight pose significant challenges due to the difficulty in distinguishing between legitimate and hostile UASs, especially in urban areas, and existing counter-UAS solutions are expensive and ineffective for real-time threat assessment.
Innovation Solution
A system utilizing machine learning to combine heterogeneous data from multiple sources for real-time anomaly detection and threat assessment, employing Object Based Production (OBP) and Activity Based Intelligence (ABI) techniques to determine the threat level of UASs, which can include crowd-sourced information and geospatial data, enabling the identification of potential threats to both conventional and non-conventional targets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional detection methods are used to identify UASs, then detection capability is limited, but the system becomes expensive and ineffective for real-time threat assessment
Solution Approach 1:
The patent combines multiple heterogeneous data sources including radar data, radio frequency data, optical data, and crowd-sourced information into a unified machine learning model. This integration allows the system to achieve reliable threat assessment by synthesizing complementary information from diverse sources, improving detection accuracy while avoiding the high costs of deploying expensive individual counter-UAS systems for each data type.
Solution Approach 2:
The patent replaces traditional mechanical and hardware-based detection systems with a software-based machine learning approach. By using algorithms that process and analyze data from multiple sources, the system achieves effective threat assessment without relying on expensive specialized hardware, thereby reducing system cost while maintaining or improving reliability.
2Measurement precision
If multiple data sources are integrated for comprehensive threat assessment, then threat level determination accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent transforms heterogeneous data from multiple sources into standardized features and parameters that can be processed by the machine learning model. By converting diverse data types (radar, RF, optical, crowd-sourced) into uniform numerical representations, the system achieves precise threat level determination while managing data processing complexity through parameter standardization and feature engineering.
3Productivity
If real-time processing of multiple data items is performed, then threat assessment speed is improved, but computational resources required increase
Solution Approach 1:
The patent implements a machine learning model that processes data items in a prioritized manner, focusing computational resources on the most critical features and data sources for threat assessment. By performing partial processing on less critical data and using efficient algorithms, the system achieves real-time threat assessment speed while reducing overall computational resource consumption compared to processing all data items with equal depth.
Data Source
AI summary
A method includes obtaining multiple data items from multiple heterogeneous data sources, where the multiple data items relate to an unmanned aerial system (UAS) flying in an area. The method also includes processing the multiple data items using a machine learning algorithm to determine a threat level of the UAS to one or more targets in the area. The method further includes outputting the threat level of the UAS for review by a user.