FP7 SUNNY Project

Smart UNattended airborne sensor Network for detection of
vessels used for cross border crime and irregular entrY
FP7 SUNNY Project
UMS 2014
William Martin
BMT Group
Exponor, Porto - 30th May 2014
The SUNNY project will develop novel and comprehensive solutions for intelligent
surveillance of maritime borders
Detection of illegal entry using a heterogeneous network of sensors carried by
Unmanned Aerial Systems (UASs)
– Smuggling
– Irregular Immigration
The surveillance of maritime borders is prioritised and sea scenarios will be used in
the project
Project Partners
Illegal incursions and border crossing
are causing social and economic
problems for EU member states;
especially Mediterranean countries
with Southern EU Maritime Borders
Strengthening border security for
immigration control and fighting
cross‐border crime is a key objective
in the EU Internal Security Strategy
2nd Quarter 2011 Frontex Risk Analysis Network Report:
• Italian islands in the Central Mediterranean identified as a hotspot for illegal
border‐crossing into the EU
• migrants from N.Africa and other sub‐Saharan countries in small sea vessels, the
majority of which in Q2 had been forcibly departed from Libya
Two-tier, intelligent heterogeneous UAS sensor network providing both
large field and focused surveillance capabilities:
First-tier sensors patrol large border
areas to provide situation awareness
and detect suspicious targets:
– Medium altitude;
– Long-endurance;
– Autonomous.
From this information, smaller secondtier sensors collect further evidence
via focused surveillance:
– track targets;
– threat evaluation;
– more accurate target recognition.
Ground station coordinates and
communicates with UAS network
Develop and integrate an airborne sensor network for the purpose of maritime
surveillance and situation awareness
– To enhance EU‐border security
The SUNNY system will implement:
– Fast, near real‐time detection and tracking of people, vehicles and ships;
– UAS mounted sensors: infrared, hyper-spectral and radar;
– On‐board, real‐time data processing;
– Adapted data link solutions from existing data link technology;
– Innovative data fusion and focus-of-attention techniques;
– An effective and intuitive human‐machine interface.
User requirements
SUNNY has questioned Navy and Customs in four Southern EU countries
– Portugal, Greece, Spain, Italy
Place most importance on preventing & detecting Drug traffic & Illegal immigration
– Then human trafficking and weapon smuggling
Wind and visibility are considered the main limitations to safety. Rain is secondary.
All groups consider UAS use important, yet most do not use any
– Preference for Small (~2h), Tactical (~3h) and Endurance (~12h)
Want to equip end-users
– Autonomous usage (no externals)
– Field training with real platforms – preferential
– Integration with older systems or partial replacement is preferred
The Imagery Intelligence (IMINT) standards were developed for observation and
surveillance using satellites and manned aircraft
– allow Imagery Analysts to detect, classify and identify objects, units or situations
The IMINT methods are:
– ”Pre-Mission Preparation”;
– “Route to Target”;
– “Data Collection”;
– ”Transfer of Primary Imagery to Exploitation”;
– ”Collect Collateral For Information Exploitation”;
– and “Exploitation” (analysis and target identification).
The methods have since evolved to account for unmanned vehicles and on-board
results exploitation – leading to a more concise set of standards for SUNNY
Threat processing
Sensors and real-time on-board
processing integrated on UAS
low weight, low cost, high resolution
that can operate under darkness, snow,
rain, etc.
Imaging sensors generate RGB,
infrared and hyper-spectral images,
and use radar information to detect,
discriminate and track objects of
Coupled sensor processing using onboard processors produce preliminary
detection results
Threat processing
Novel identification algorithms,
independent of target and terrain
knowledge, analyse UAS data;
e.g. using deviations of colours, shapes
and textures between suspicious
objects and their surroundings
UAS information is combined with:
– target models;
– other data, e.g. AIS;
– environment knowledge;
– global situation;
– impact assessment.
A 3D visualisation UI will show:
– assets (e.g. borders):
– sensors (e.g. UASs);
– agents (e.g. vessels).
Data compression
All UAS activities will use civilian frequencies and protocols on an adapted data-link
– limitations in bandwidth / range
Raw UAS sensor data cannot be concisely represented for human operators
Human operator assisted decisions will require small timescales
– data must be filtered to a logical format reducing non-relevant information
• Automatic data pre‐processing and compression on‐board UASs achieves data
reduction between the vehicle and ground stations
– compression must not reduce accuracy or image resolution;
– the proper selection of relevant information and images must occur
Sensors summary
Developing multi‐band IR cameras for agile platform use
– Low power consumption with a high temperature range
Hyper-spectral sensors utilise characteristic spectral signatures
– includes information about fine colour and chemical differences
– on board data processing is mandatory to reduce the very high raw data output
Developing compact, low power CW radar technologies for both UAS tiers
– image processing will compensate for image blur due to relative motion of target
– 3D radar imaging will enable more robust object recognition than existing radar
Example scenario for the SUNNY system:
Tier 1 UAS detects a target vessel using Long-Wavelength Infrared sensors
• Target detection supported by CW radar sensors
Ground control validates target and sends in Tier 2 UASs
Tier 2 Hyper-spectral (VisNIR ) sensors can identify the number of persons on-board
Visibility degrades between Tier 1 detection and Tier 2 deployment:
• CW radar sensors allow continued operation
• Thermal LWIR hyper-spectral sensors can enable night-time operation
Threat evaluation is made – relevant authorities are notified to take action
SUNNY intends to develop a new tool for collecting real‐time information in
operational maritime scenarios
SUNNY represents an advancement due to the following features:
– A two‐tier, intelligent UAS sensor network;
– Novel sensors and on‐board data processing, integrated on UAS systems;
– The exploitation and adaptation of emerging standard wireless technologies and
SUNNY will deliver pre‐processed information with meaningful decision support tools
– to enhance and support the capabilities of border surveillance personnel
"The use of UAVs in the framework of EUROSUR is excluded for the
time being.
Their potential future use depends on a number of pre-conditions,
such as their full integration into civil/non-segregated airspace
(planned for 2016 at the earliest) and further operational testing,
showing that this technology is at least as cost -effective and efficient
as manned surveillance planes (which currently is not the case).
This means that any assessment whether the results of the FP7
SUNNY project can be used for border surveillance respectively in the
framework of EUROSUR can only be carried out after the SUNNY
project has been finished in 2017."
Infrared sensors
Developing multi‐band IR cameras for agile platforms
– Long-Wavelength Infrared (LWIR) information for detection;
– Visible Near (VisNIR) and Short-Wavelength Infrared (SWIR) for identification.
Low power consumption with a high temperature range
LWIR uncooled IR camera engines are already used in static surveillance
– advantages of lower weight and less energy consumption
Integration of IR camera with UAS still requires:
– mechanical and electrical integration,
– communication and data link connections,
– data processing, analysis and visualization connections
(VisNIR: 600-1600 nm, LWIR: 8-15 µm, SWIR: 1.4-3.0 µm)
Hyper-spectral detection
Hyper-spectral (HS) imaging is uniquely able to both detect targets and to identify
them, by using their characteristic spectral signatures
– Improvement over traditional RGB and broadband thermal imaging
Spectral signatures include information about small colour differences and particularly
about chemical differences
– VisNIR region can discriminate between vegetation and man-made targets in
day-time operations;
– Thermal LWIR enables night-time operation.
On board data processing is mandatory to reduce the very high raw data output
– Integration of multiple-core GPU
HS imagery fusion, from multiple UASs, to generate wider, improved situational
Radar sensors
Innovative CW miniSAR sensor for both UAS tiers
Features new signal and radar imaging processing for enhanced target identification
– will extend monitoring operations in reduced visibility conditions;
– contributes to the detection and identification phases.
Developing compact, low power CW radar technologies
Innovative ISAR image processing to compensate for image blur due to relative
motion of target
Developing innovative 3D radar imaging, enabling more robust object recognition
than existing radar sensors
(SAR = Synthetic Aperture Radar; ISAR = Inverse-SAR; CW = Continuous Wave)

similar documents