RSS feed Add to your Facebook page Watch us on Youtube

Activity title

ARTIFICIAL INTELLIGENCE FOR MILITARY MULTISENSOR FUSION ENGINES

Activity Reference

SET-262

Panel

Sensors & Electronics Technology

Security Classification

NATO UNCLASSIFIED

Status

Awaiting Publication

Activity type

RSM

Start date

2018

End date

2018

Keywords

anomaly detection and sensor data integrity, examples from military applications, machine learning and Bayesian methods for sensor data exploitation, multiple hypothesis and logical analysis, probabilistic reasoning and decision making with big and sparse sensor data, sensor and resources management, tracking and multiple sensor data fusion

Background

I. BACKGROUND: Mathematical methods of Artificial Intelligence AI may substantially improve MULTISENSOR FUSION ENGINES that are already and will be even more so the backbones of NATO’s situational awareness capabilities with disruptive effects for friendly and adversary forces. FUSION ENGINES provide militarily relevant information from heterogeneous multi-functional multiple sensors on multiple distributed and moving platforms that: • have to be embedded into overarching C5(J)ISR systems; • have to profit from advanced communications links; • have to be cyber-safe and classically EW resistant; • have to be integrated into manned-unmanned teaming; • have to be embedded into intuitive human machine interfaces. The proposed RSM is a sensor-focussed counterpart of the planned 2018 STB Specialists Meeting “Decision Making using Big Data and AI” proposed by Michael Wunder, IST Panel Chair. It will discuss specialised AI-inspired algorithms for multiple sensor data exploitation and sensor resources management. Moreover, it is related to the proposed RLS “Artificial Intelligence for Military Multiple Fusion Engines”, to be held in Budapest (HUN), Wachtberg (DEU), Atlanta, San Diego (USA) which will be sharing knowledge on military sensor informatics in a tutorial way. The proposed RSM will profit from previous STO activities such as SET-218 RTG on “Interoperability & Networking of Disparate Sensors and Platforms for ISR Applications”, SET-235 LS on “Radar and SAR Systems for Airborne and Space-based Surveillance and Reconnaissance”, IST-155 LS on “Advanced Algorithms for Effectively Fusing Hard and Soft Information”.

Objectives

The scientific objective of the proposed RSM is to share latest research and expertise on sophisticated AI inspired sensor informatics of among the NATO’s member states. An equally important objective is to encourage further R&D on sensor-focussed Artificial Intelligence that is addressing the needs of NATO’s future missions. Moreover, an effective network among the AI stake holder in NATO’s member states will be a natural result, which is a third objective of the proposed RSM. The RSM will present invited papers and presentation on advanced methodologies and algorithms of AI inspired sensor informatics that may have disruptive effects on various aspects in the context of military situational awareness. To be encouraged will be papers on advanced artificial intelligence methods and sensor-related examples taken from: probabilistic reasoning over time, statistical decision making, big/tall/sparse data in multiple sensor data fusion for tracking, classification, anomaly detection, Bayesian and machine learning methods, knowledge-representation, multiple hypothesis and logical analysis, sensor and resources management, examples from military applications. Expected deliverables will be RSM proceedings with all invited papers and presentation slides. Since MILITARY MULTISENSOR FUSION ENGINES are already in military use, progress achieved by the RSM may have an rather direct impact.

Topics

Needless to say, sensor-focussed Artificial Intelligence is a “hot topic” in the civilian world as well. With emphasis of military applications of AI methods, the RSM will discuss latest research results on sensor AI topics such multiple hypothesis analysis, advanced multiple object characterization, distributed decision making, learning algorithms for object, situation and mission features etc. Multi-functionality is and will be a predominant factor, i.e. the shared use of sensing hardware to achieve specialized goals. Emphasis will also be placed on guaranteeing sensor data integrity and security aspects of MILITARY MULTISENSOR FUSION ENGINES. Besides classical Electronic Warfare issues, this comprises “malicious” or “byzantine” sensors as well as navigation and cyber warfare. Moreover, tomorrow’s MILITARY MULTISENSOR FUSION SYSTEMS will be inherently adaptive to scenario and mission requirements and will exploit external knowledge bases massively, which will be another issue to be covered.

Contact Panel Office