Aerospace and Electronic Systems Magazine September 2016 - 19

also execute avoidance maneuvers. The different kinds of collision
avoidance solutions form a layered approach, which is illustrated
in Figure 1.
The authors' work is focused on the SAA layer as the testbed,
and targeted aircraft category pertains to small or mid-size UAVs,
where the physical constraints of the vehicle and limited by budget
constraints, which make the integration of any other kind of systems (ACAS for example) unviable.
In the literature there are many approaches to address the SAA
problem, with different kinds of aircraft, sensors, and environment
conditions in which the task is run. In [5], several sensor technologies were examined to determine which would be a good candidate
for the main sensor of a UAS SAA system. The tested sensors are:
electro-optical (EO) and infrared (IR) camera, microwave RAdio
Detection And Ranging (RADAR), Laser RADAR (LADAR), and
bistatic RADAR. Although the EO camera had the best score among
them, the LASER and the microwave RADAR had similar performance. However, the LASER and microwave RADAR can easily
violate the weight, size, and power constraints of a small UAS. For
these reasons, this work considers vision-based approaches, which
can be possibly applied even on the smallest UAVs.
One of the best current SAA solutions in this field is introduced
in [6]. The referenced system uses information from RADAR, as
well as from an EO camera. This system consists of two IR cameras, two regular EO cameras, and a RADAR next to the conventional guidance navigation and control system. Additionally, it was
developed for a High-Altitude Long-Endurance UAV. The size and
weight of this particular UAV was comparable to a lightweight
commercial aircraft. The system was tested on a Tecnam P92 with
a wingspan of 8.7 m and a weight of 450 kg. Accardo et al. showed
that the RADAR used was capable of detecting the intruder aircraft
reliably [6]. The ranges were compared to Global Positioning System measurements. It was shown that in low altitudes there was
significant noise in the radar-based detection due to ground clutter.
The system provided reliable situational awareness with a 10 Hz
update frequency. However, it is pointed out that misalignment between the camera and radar sensors can lead to inaccuracies in the
fused detection system.
The main advantage of the proposed system is that it is capable
of running the SAA in all-time, all-weather conditions. Due to the
SEPTEMBER 2016

camera sensor, its direction estimate is more reliable and more accurate than other RADAR systems. The main drawbacks of the
system are: the problem caused by the fusion of different sensors;
the cost of the system is expensive because of the high number of
sensors used; and it is heavy as well, so it cannot be used on a midsize or small UAVs.
Another promising concept was shown in [7]. This system was
developed at the Queensland University of Technology, Australia,
as a part of the Australian Research Centre for Aerospace Automation's Smart Skies research project. The main advantage of this
research project is that they have access to various types of aircraft,
sensors, and computational resources, and have a big database of
flight videos collected in various situations. The detection is based
on a Close-Minus-Open morphological filter and a new Hidden
Markov Model (HMM) temporal filtering method, with the addition of relative bearing and elevation estimation capabilities. They
have shown an extensive evaluation for the system in their previous publications with the description of a novel collection methodology for collecting realistic airborne collision-course target footage in both head-on and tail-chase engagement geometries.
The morphological-HMM-based approach was shown to be
able to achieve reasonable detection ranges at very low falsealarm rates (in both blue sky and cloudy conditions). The detection
ranges and false alarm rates are very impressive, and the authors
have the biggest known airborne video database as well, with a real
target aircraft.
The main drawback seems to be the power consumption of
the proposed system due to the computationally extensive preprocessing and temporal filtering steps. The authors built a Graphics
Processing Unit (GPU)-based system for the detection. The power
consumption of the GPU itself is 59 W, and there is a host computer next to it, which seems to be too much for a small size UAV.

Figure 1.

The layered collision avoidance concept.

IEEE A&E SYSTEMS MAGAZINE

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine September 2016