RSS feed Add to your Facebook page Watch us on Youtube

Activity title

Employing AI to Federate Sensors in Joint Settings

Activity Reference

SAS-ET-EI

Panel

System Analysis and Studies

Security Classification

NATO UNCLASSIFIED

Status

Active

Activity type

ET

Start date

2019

End date

2020

Keywords

Artificial Intelligence, JISR, Sensors

Background

Surveillance of an area of interest is often performed by several distinct sensor capabilities, each with its own closed target identification system. Through the JISR concept, NATO is aiming to collect, process and disseminate target information from different systems to relevant users. Although the target identification may be very advanced, e.g. AGS with its Automatic Target Identification System using machine learning to continuously improve its success rate, the identification is based on one type of sensor. Validation of target identification based on different types of sensors are left until late in the intelligence cycle. It might be more effective to first evaluate the physical characteristics of the scenario and the expected types of targets, and then, based on this analysis, to specify the best mix of sensor types and number of sensors to obtain the highest possible success rate from a target identification system using unprocessed data from the selected sensors. The activity should take into account that traditional sensor systems may be augmented by swarms of unmanned ISR vehicles, as being investigated in SET-263. This activity will focus on selecting types and numbers of sensors to federate.

Objectives

The objective of the activity is to be able to identify unique characteristics of different kind of targets enabling operational planners to design the most effective federated network of available sensor systems including autonomous unmanned ISR vehicles in a given operational scenario.

Topics

• Characterization of operational scenarios based on their physical characteristics. • Detectability of different kinds of targets in different kinds of physical conditions. • Methods to set-up and manage an ISR swarm in a given scenario. • Probability of detection as a function of types and numbers of sensor feeds. • Probability of false positives as a function of types and numbers of sensor feeds.

Contact Panel Office