SPECIAL REPORT:

UNMANNED SYSTEMS
MARCH 2020

Sponsored by
Your Stamping &
Washer Specialists
REQUEST
YOUR FREE
STAMPINGS
& WASHERS
CATALOG
& CALENDAR!

3D-Printed
Prototypes
Available

bokers.com
(888)-WASHERS 612-729-9365 sales@bokers.com
CONTENTS
FEATURES
2 Artificial Intelligence and
Autonomous Vehicles
6 Underwater Drone Technology
10 The CAN Bus: Driving the
Future of Autonomous
Military Vehicles
14 Autonomous "Wingman"
Vehicles
18 Microturbine Propulsion
for UAVs
20 Eyes in the Sky
APPLICATION BRIEFS
22 
Parcel Delivery Drone
23 Robotic Helicopter
UNMANNED SYSTEMS SPECIAL REPORT
TECH BRIEFS
24 Aircraft Vertical Takeoff and Landing
24 N
 eural Lander Uses AI to Land Drones
Smoothly
25 Lightweight Sensing and Control System
for Unmanned Aerial Vehicle Monitoring
25 Flying Robot Mimics Rapid Insect Flight
26 Software Enables Robot to Self-Right
After Overturning
ON THE COVER
Seventy years ago, military aviation
moved from reciprocating engines to
vastly more reliable turbo jets and
turboprops. The same cannot be
said for UAVs. The industry has been
slow to innovate and develop turbine
propulsion systems for small aircraft
because it is far more difficult to
produce high-performance small
turbines than large ones. But that
may be about to change. See the
article on page 18.
(Photo by sibsky2016/Shutterstock.com)
MARCH 2020 1
ARTIFICIAL
INTELLIGENCE AND
AUTONOMOUS VEHICLES
T
he use of artificial intelligence (AI) based machine
learning technologies in autonomous vehicles is on
the rise. Helping to drive this trend is the availability
of a new class of embedded AI processors. A good
example is NVIDIA’s Jetson family, which includes small
form factor system on modules (SoMs) that provide GPU-
accelerated parallel processing. These high-performance,
low-power devices are designed to support the deep
learning and computer vision capabilities needed to build
software-defined autonomous machines. They derive massive
computing capabilities from the use of a parallel processing
GPU device with many cores, enabling next-gen computing
devices to take on many of the tasks that were historically
handled by humans or multiple, traditional computers.
How AI provides navigation and obstacle avoidance
for autonomous vehicles on land, air, and sea gets a lot
of attention, but machine learning is also being used in
other ways on unmanned vehicles. These machine learning
applications are usually related to the types of sensors that
are onboard a particular platform. For example, a number of
our customers are using AI engines in unmanned situational
awareness airborne applications, taking and processing
data from different types of sensors, such as cameras,
radars, and LiDARs, which are used for surveillance and
detection, and then report that data back to operators.
2 MARCH 2020
The ability of unmanned aircraft to monitor areas
of interest, staying up in the air for long periods of
time, makes them a compelling alternative to the use
of human crews, especially in battlefield situations.
In one application, a homeland security agency is
protecting critical national energy infrastructure by
using drones to monitor oil pipelines. In addition to
surveillance applications, AI engines are now being
used on unmanned air vehicles for in-vehicle health
monitoring. Sensors onboard the platform are monitored
by a predictive maintenance data hub that uses AI to
monitor the health of various onboard subsystems, which
helps to schedule maintenance and other updates.
The available market for military use of AI engines is
relatively small compared to the industrial applications
for which most of these devices were originally intended.
The larger opportunities for vendors of GPU-based AI
engines are likely found in places like factory floors and
production lines. That said, these devices are increasingly
used in the commercial vehicle market onboard autonomous
test cars. For military applications, the use of these
devices on unmanned platforms deployed in harsh
environmental conditions requires expertise in packaging
and ruggedization to ensure that devices built for industrial
use can perform optimally when deployed in the battlefield.
UNMANNED SYSTEMS SPECIAL REPORT

The NVIDIA Jetson TX2i module provides a great example
of how an industrial AI engine can be adapted for use
onboard military platforms, most of which are especially
sensitive to any additional size, weight, and power (SWaP)
burden. Another key issue for defense applications is long
lifecycle support, since commercial graphics processors
are notorious for short product lifecycles. The good news
for military system designers is that SOMs like the Jetson
offer some of the best ratios of FLOPS of computing
performance per watts of power consumed. The TX2i,
specifically, combines a high-performance 6-core Arm
processor with a powerful 256-core NVIDIA GPU that is
compatible with their CUDA development libraries and
software applications. Optimized to reduce SWaP, this
device is packaged in a small form factor pre-integrated
with processor, graphics, RAM memory, Flash storage, and
I/O. With its support for extended operating temperatures
(all of its components are rated to the full industrial
temperature range of -40°C to +85°C or beyond), the
TX2i can be effectively cooled using straightforward
conduction cooling, eliminating the need for complex cooling
alternatives or fans that have undesirable moving parts.
For enhanced reliability, the module uses RAM memory
with Error Corrected Code (ECC) support, which is
particularly useful to mitigate a potential single event
upset (SEU) at high altitude due to solar radiation. Even
better, NVIDIA supports the industrial version of this
Jetson device with a ten-year lifecycle, which matches
the long program requirements typical of military
systems. Compare that to the five-year lifecycle of other
embedded Jetson devices and an even shorter two-
year lifecycle for consumer-grade graphics devices.
To ease and speed application software development, the
TX2i is supported by NVIDIA’s JetPack Software Development
Kit (SDK) that includes a large number of free CUDA
UNMANNED SYSTEMS SPECIAL REPORT
NVIDIA’s Jetson TX2i small
form factor system on
module (SoM).
developer tools and comprehensive libraries for building AI
applications. TensorRT, a deep learning inference runtime,
and the CUDA development environment are just a couple
of free tools bundled in JetPack. If the system integrator
is already familiar with the NVIDIA Jetson development
environment, they will also be comfortable with this
technology, and since it’s portable across the entire Jetson
family ecosystem, the software development environment
is portable across all Jetson modules, reducing costs
through a build-once and deploy multiple times approach.
NVIDIA also offers lab-grade developer kits at an
extremely low cost. Designed to jumpstart application
development, these boards use standard commercial
connectors and enable customers to get their program
started with very little investment in the lab before
porting it over to a rugged platform when they are
ready to take the next step towards deployment.
An example of a rugged mission computer ideal for use in
deployed machine learning applications is Curtiss-Wright’s
DuraCOR 312, a miniature system (5.2" x 5.4" x 2.0") based on
the industrial Jetson TX2i SoM. The DuraCOR 312 can be used
in a wide range of AI-based applications and is especially
useful for computer vision applications. For example, the
mission computer can be deployed on a small drone and
flown around a platform of interest to inspect for damage
or any visible anomalies. At sea, the system can be used
to locate and identify surface vessels. In the battlefield, a
DuraCOR 312 can be launched on a small unmanned aircraft
to provide object detection and visual detection, providing
real-time actionable intelligence of enemy troop strength
and deployment without needing to request and wait for air
support, such as a large helicopter, to provide surveillance.
In the case of visual detection, the video captured by
the unmanned aircraft’s camera is sent to the GPU, which
runs a machine learning inference, basically scanning the
MARCH 2020 3
ARTIFICIAL
INTELLIGENCE AND
AUTONOMOUS VEHICLES
Curtiss-Wright’s
DuraCOR 312 miniature
system based on NVIDIA’s
Industrial Jetson TX2i SoM.
images to identify, as it’s been trained to do, any objects
of interest, such as a soldier or tank, and it will tag the
detected objects with time and location data. The inference
code, which is stored on the DuraCOR 312, was developed
earlier on a larger system using an algorithm of choice.
Images associated with the objects that you want to detect
are formatted appropriately and then used to train the system
to recognize them. After the system is trained satisfactorily
to the desired percentage of accuracy, the machine-learning
model, written using a deep learning framework such as
Caffe or TensorFlow, is then run in TensorRT, which helps
to compress the software code for use on small systems
deployed at the edge of the battlefield network. The inference
code, once downloaded onto the DuraCOR 312, can now
be deployed on a drone and begin to process incoming
images to identify those objects it was trained to recognize.
The results can then be downlinked to the warfighter in
real time or stored on the drone until it returns for offline
post-mission analysis. How much storage is required on the
aircraft will be determined by the distance and duration
of the mission and the amount of imagery captured.
When deployed in battlefield environments, cooling the
TX2i’s power-efficient ARMv8 processor cores and CUDA
GPU is aided through the use of an aluminum heat spreader
on top of the module that makes contact with all of the
hot components. The heat spreader attaches directly to
the DuraCOR 312’s system chassis, using phase change
material, and the metal of the chassis as an extension of the
heatsink. To further adapt the TX2i for use in rugged military
conditions, Curtiss-Wright submits all system components
to additional ruggedization. The DuraCOR 312 also uses
miniature versions of traditional rugged circular military
connectors that, when mated, provide full sealing against
4 MARCH 2020
dust and water. In fact, the chassis can be fully immersed
and will experience no water ingress at all. What’s more, all
the circuit boards are covered with a conformal coating to
protect against humidity and corrosion. Designed for SWaP
optimization, this compact AI-engine-based mission computer
weighs less than 2.0 lbs. and requires less than 25W of power.
The DuraCOR 312 is fully compliant to extremely
demanding MIL-STD-810G, MIL-STD-461F, MIL-STD-
1275D, MIL-STD-704F, and RTCA/DO-160G environmental,
power, and EMI conditions, including high altitude,
wide temperature, humidity, extreme shock, and
vibration and noisy electrical environments. The unit
also provides an aerospace-grade power supply in a
fanless IP67-rated mechanical package that handles
harsh shock and vibration and operates over extended
temperatures without requiring a cold plate or airflow.
Through our experience collaborating with companies like
Cisco Systems to deliver rugged network switch solutions
for challenging embedded applications, we’ve learned the
benefits, both for the customer and for the COTS vendor,
that come from partnering with a technology pacesetter.
Now, by leveraging NVIDIA’s best-in-class industrial AI
engines, and packaging them for the military COTS user, we
can bring proven, deep learning technologies, supported
with a rich and familiar development environment, that
are viable for the unique requirements of unmanned
military platforms. This approach reduces program risk
and provides system designers with the shortest and
easiest learning curve to shorten time to deployment.
This article was written by Mike Southworth, senior
product manager, Curtiss-Wright Defense Solutions
(Salt Lake City, UT). For more information, visit
http://info.hotims.com/72998-502.
UNMANNED SYSTEMS SPECIAL REPORT
APPLICATION-SPECIFIC INTERCONNECTS FOR DEFENSE SOLUTIONS
WWW.OMNETICS.COM | SALES@OMNETICS.COM | +1 763-572-0656

OMNETICS IS A WORLD-CLASS
MINIATURE CONNECTOR DESIGN AND
MANUFACTURING COMPANY WITH
OVER 30 YEARS OF EXPERIENCE. OUR
MINIATURE CONNECTORS ARE
DESIGNED AND ASSEMBLED IN A
SINGLE LOCATION AT OUR PLANT IN
MINNEAPOLIS, MINNESOTA.

WE TAKE PRIDE IN WHAT WE

BUILD FOR YOU.
Underwater Drone
Technology
F
inding ways to overcome
physical limitations so that
humans can dive deeper
and stay underwater longer
has been an ongoing quest. Back
in the 15th century, Leonardo da
Vinci drew sketches of a submarine
and a robot. Had he thought to
combine the two concepts, he would
have created a prototype of an
unmanned underwater vehicle, or
underwater drone. Instead, the world
had to wait another 500 years.
In the 1950s, in response to sea
mines that would go on to destroy or
damage more naval ships than all other
types of attacks, the U.S. Navy began
developing an unmanned underwater
vehicle (UUV) that could dive 10,000
feet (rated depth) and function for
four straight hours. In the succeeding
decades, the Navy developed UUVs that
could retrieve lost equipment such as
torpedoes or a missing atomic bomb,
REMUS 6000
6 MARCH 2020
save submarine crews lost at sea, or
locate shipwrecks, including the Titanic.
At the beginning of the 21st century,
UUVs increasingly became a faster,
cheaper, and safer alternative than
sending humans and trained animals
to defuse sea mines. For example, it
takes a team of human divers 21 days
to clear sea mines in one square mile,
but a UUV such as the Hydroid REMUS
can complete the task in only 16 hours.
How Do Unmanned Underwater
Vehicles (UUVs) Work?
Unmanned underwater vehicles
consist of remotely operated vehicles
(ROVs) and autonomous underwater
vehicles (AUVs).
Remotely Operated
Vehicles (ROVs)
ROVs were developed first and are
controlled by humans from a distance.
Their sizes vary.
All ROVs include built-in sensors, a
tether, a tether management system,
a flotation pack, and a thruster. Most
ROVs have sensors for video cameras
and lights. They may also have sonars,
magnetometers, water samplers,
and instruments that measure water
clarity, water temperature, water
density, and sound velocity. An ROV
might also include a 3D camera
to improve the pilot’s perception
of the underwater scenario.
Linked to a host ship by an umbilical
tether that houses energy cables
and communication cables, ROVs
use a tether management system
(TMS). The TMS makes it possible
to adjust the length of the tether to
minimize the underwater effect of
cable drag. By employing a flotation
pack and a thruster, heavy-duty
ROVs can achieve precise motion
control even in high-current waters.
HYDROID
UNMANNED SYSTEMS SPECIAL REPORT
 A-size autonomous
underwater drone.

General
Dynamics Boeing Lockheed Martin Teledyne Saab Kongsberg Hydroid
(Bluefin)

Product Bluefin-21 Echo Voyager Marlin Gavia Sabertooth HUGIN 1000 REMUS 600
example

Type AUV AUV AUV AUV AUV/ROV hybrid

Weight 1650 lbs. 100,000 lbs. 3,500 lbs. 105 lbs. 1400 lbs. 1400 lbs. 500 to 850 lbs.

Rated Depth 15,000 ft. 11,000 ft. 12,000 ft. 3000 ft. 3600 ft. 9000 ft. 1800 ft.

Power source Lithium Battery, rechargeable, Battery, Battery, Battery, Battery, Lithium battery,
battery, and marine diesel rechargeable rechargeable rechargeable rechargeable rechargeable
rechargeable generator

Estimated 25 hours at 3 months; or 60 hours over 14 7 hours 8 hours 100 hours at 4 Up to 24 hours
Endurance 3 knots 280 km days; 100 km knots

Navigation Inertial Inertial Navigation, Doppler aided Inertial IMU/Doppler and Inertial Inertial
navigation Doppler Velocity inertial navigation, navigation, terrain naviga- navigation, navigation,
Log, depth sensors USBL, terrain Doppler velocity tion micro-navigation acoustic doppler,
navigation meter, USBL sonar

Application Research, Oil and gas, Oil and gas, Research, Oil and gas, Research, oil and Research,
military military military military military gas, military military

Table 1: Comparison of underwater drones (not all models are included).

Smaller ROVs are used mainly for
observation and inspection in the
research, military, or recreational
sectors because they can only
withstand shallow deployment (around
1,000 feet or 300 meters) and have
limited horsepower. On the other
hand, the largest and most heavy-
duty ROVs are suitable for the oil and
gas industry because they can dive
as deep as 20,000 feet and perform
multiple tasks like lifting, drilling,
construction, and pipeline inspection.
Autonomous Underwater
LOCKHEED MARTIN
control and do not have to be tethered
to a larger vessel. Having exchanged
the tether for more mobility and
freedom, AUVs must face the problems
of communicating with the ships,
navigating on their own, and powering
themselves over a long period.
Communication: Water distorts
signals, so the usual means of
communication including GPS, mobile
(3G, 4G), Wi-Fi, radar signals, and
optical navigation such as LiDAR will all
be ineffective for AUVs. While useful,
very low frequency (VLF) and extremely
low frequency (ELF) signals can only
low bandwidth (1-2 kbps) that restricts
the rate of data transfer. Acoustics
may therefore be the most suitable
communication solution for AUVs, but
challenges include signal diffusion,
relatively low bandwidth, and power
consumption. At 20 kbps, the bandwidth
offered by acoustic modems is still
insufficient for real-time video, meaning
that the human operator on the ship
will only see delayed images. The use of
high-definition cameras may supplement
the collection of information.
Navigation: Signal distortion by
water also hampers how well AUVs

Vehicles (AUVs)
In recent years, military applications
have taken the next step and begun
using AUVs, which, unlike ROVs,
operate independently of human
enable one-way communication from
the ship to the drone, since VLF or ELF
transmitters are too bulky to mount
on the drone. Also, VLF and ELF have
long wavelengths and thus extremely
orient themselves. Therefore, most
AUVs use inertial navigation. Inertial
navigation supplements dead reckoning
calculation of the drone’s current
position based on the previously

UNMANNED SYSTEMS SPECIAL REPORT MARCH 2020 7

Underwater Drone
Technology
determined position and known or
estimated speed over time and course
with motion sensor (accelerometer)
and rotation sensor (gyroscope) data.
Also, AUVs use a Doppler velocity
meter that acoustically estimates
the drone’s velocity relative to the
sea bottom and ultra-short baseline
(USBL) positioning, which works by
emitting a regular acoustic signal from
a surface reference point, such as a
support ship’s underwater beacon.
Power: AUVs must run reliably and
safely for days, even weeks, without
human intervention. Therefore, they
need more efficient ways to move
around or consume energy. For
example, by converting buoyant,
vertical motion to horizontal motion
with its wings, an underwater glider
propels itself forward with little energy.
While most AUVs use rechargeable
lithium batteries, some of the larger
AUVs use aluminum-based semi-fuel
cells that are higher-maintenance.
Independence: Researchers are
also trying to make AUVs more
independent by developing machine
intelligence solutions that enable
AUVs to navigate complex situations
and solve problems by themselves.
Applications
With stronger processing power
and a bigger power supply, AUVs
can take on more challenging
projects, such as locating the sunken
Argentine navy submarine ARA
San Juan in 2018 and recovering
the black boxes from the crashed
Air France Flight AF447. AUVs are
also poised to stretch the range of
applications they take on. Used for
ship hull inspection, wreck inspection,
nuclear reactor decontamination,
exploration, mining, drilling, and
wreck recovery, they have caught the
attention of researchers investigating
their capabilities for weather
forecasting, hurricane hunting,
and environmental conservation.
Among the major players in the
AUV space (see Table 1), several
of them have especially close
ties with the U.S. military.
8 MARCH 2020
The REMUS drones by Hydroid
probably have the oldest and most
recognizable brand, mainly due to high-
profile operations involving REMUS
drones such as sea mine detection
during Operation Iraqi Freedom. Also,
Canada, Japan, and NATO countries
are using REMUS AUVs in their military.
Of the many types of REMUS drones,
REMUS 6000 is the biggest unit.
General Dynamics’ Bluefin-21 is
most famous for its involvement in
the search for the wreckage of the
missing Malaysia Airlines Flight 370
in 2014. The U.S. Navy had previously
ordered several Bluefin-21 UUVs
for its Black Pearl AUV program.
Currently, the U.S. Navy is looking to
improve the Bluefin-21, especially its
two-way communication capability.
Meanwhile, General Dynamics is
developing the Knifefish, which is
based on Bluefin-21 and has improved
capabilities in detecting mines and
working with Littoral combat ships.
Boeing’s Echo Voyager is, by far, the
biggest AUV around (hence dubbed
“XLUUV”). At close to 100,000 lbs., it
has the longest endurance and one of
the deepest rated depths. Echo Voyager
is also completely autonomous, as it
does not have to be launched from,
controlled by, or recovered to a
support vessel. What’s more, Boeing
is building the Orca, which is based on
Echo Voyager and promises to have
more warfare capabilities than just
those of mine removal so that it can
potentially substitute for the much
more expensive Littoral combat ships.
Lastly, Boeing has been acquiring AUV
start-ups to expand its portfolio.
Lockheed Martin is another
contractor making extra-large drones,
which in this case are the Marlins.
Marlins also have extended endurance,
range, power, and rated depth. In
addition, the Marlin can generate real-
time, 3D, geo-referenced models of
surrounding environments from the
data it collects. Lockheed Martin is
also maintaining a presence in the
small AUV space with its line of A-size
drones that act as training targets
for submarines, aircraft, and ship
crews. Like Boeing, Lockheed Martin
is also acquiring AUV start-ups as
part of its expansion in the sector.
Challenges and Future
Perspective
Manufacturers aim to continually
improve the functionality, performance,
eco-friendliness and power efficiency of
their UUVs. Saab has already launched
an AUV/ROV hybrid that can switch
between roles during a mission, and
there will be more hybrid launches in
the future. Also, the sector is turning to
biomimicry to improve propulsion and
maneuverability, as in the case of Festo’s
AquaJelly and EvoLogics’ BOSS Manta
Ray. Moreover, there is concern that
the USBL technology disrupts marine
life and could interfere, for example,
with the acoustic communications
whales use when feeding, breeding,
and breaching. Lastly, researchers
are experimenting with combining
different battery and power systems
with supercapacitors, extracting energy
from seawater, installing underwater
power stations to refuel the UUV,
and transferring power wirelessly.
Manufacturers are likely responding
to the continual commitment from
the government. Within the U.S.
Department of Defense, the Navy
receives the biggest budget to develop
unmanned systems. The Navy plans
to have its squadron of UUVs by
2020 and ultimately replace trained
animals with Kongsberg/Hydroid’s
REMUS MK 18 Mod 2 Kingfish in
sea mine removal operations.
Meanwhile, navies in the UK, France,
Russia, Japan, and China continue
to use UUVs or develop related
technologies to stay competitive. For
example, China is working to incorporate
artificial intelligence into drones, and
Russia is developing nuclear-powered
drones. Whether for the purposes of
exploration, research, rescue, military, or
defense, one thing is clear — advanced
technologies used in underwater
drones will continue to develop.
This article was written by John
W. Koon, contributing editor,
Aerospace & Defense Technology.
UNMANNED SYSTEMS SPECIAL REPORT
Cyber Defense
D
AT THE

Tactica
a al
Ed
Edge

RUGGED. SCALABLE. SECURE.

Achieve accreditation fo
for fo
forward-deployyed
programs — with ease.

Combat-proven solutions from Crystal Group protect

critical data and eliminate guesswork. As a trusted
partner on more than 600 U.S. deffeense programs, our
high-perfo
formance rugged hardware is engineered with
certified encr yption and anti-tamper fe
features to address
complex cyber requirements.

When lives are on the line, failure is not an option.

Crystal Group delivers. Learn how our solutions integrate
demanding perfo
formance and security standards fo
for
accreditation of your program.

SERVERS | DISPLAY
AYS | STORAGE | NETWORKING
G | EMBEDDED | CYBER
sales@crystalrugged.com | 800.378.1636 | crystalrugged.com
 The CAN Bus:
DRIVING THE FUTURE OF AUTONOMOUS
MILITARY VEHICLES
I
t’s a crisp November day in Michigan,
and a convoy of British and American
resupply vehicles is rumbling along
at a comfortable 25 miles per hour. In
the lead is a British Army Rheinmetall
MAN Military Vehicles (RMMV) HX-
60 truck, trailed closely by two U.S.
Army Oshkosh Light Medium Tactical
Vehicles (LMTVs). In total, there are
zero humans operating this convoy.
This was the scene in November
2017, when the US-UK Coalition
Assured Autonomous Resupply
(CAAR) demonstration took place
at Camp Grayling, Michigan — the
result of three years of research and
engineering to examine the use of
unmanned ground vehicles (UGVs)
in resupply situations to tackle so-
called “last mile” issues regarding
supplying troops on the front lines of
active conflicts around the world.
CAAR is a collaboration among the
United Kingdom’s Defence Science
and Technology Laboratory (Dstl); the
10 MARCH 2020
U.S. Army’s Tank Automotive Research,
Development, and Engineering
Center (TARDEC); and the U.S. Army
Armament Research, Development, and
Engineering Center (ARDEC). While
this project had been in the works
for three years, the U.S. military has
been exploring autonomous vehicle
technology in earnest since at least
the 1980s. In the time period since
then, there have been huge leaps
in autonomous vehicle technology,
and the military has seen already
dangerous and logistically difficult
resupply issues multiplied in the
CAN Bus Board
era of improvised explosive devices
(IEDs) and protracted conflicts in
remote and desolate terrains.
The technologies powering these
UGVs are cutting-edge. Many could
hardly have been envisioned a decade
ago, but the secret weapon of nearly
all of these vehicles is a technology
that has been standard in civilian
vehicles for nearly 30 years: the
Controller Area Network, or CAN bus. BACKGROUND IMAGE: FABER14/SHUTTERSTOCK.COM
What is a Controller Area
Network (CAN)?
A controller area network (CAN) is
a message-based protocol that allows
internal systems to communicate with
one another without a central computer.
CAN technology is used in applications
as wide-ranging as agriculture, robotics,
industrial automation, and medical
systems, but it is most known for its
use in the automotive industry.
In today’s connected vehicles, the
CAN bus facilitates communication
UNMANNED SYSTEMS SPECIAL REPORT
between microcontrollers (MCUs)
along a larger vehicle bus, without
the use of a central computer. For
example, the cruise control system
can quickly communicate with the
anti-lock braking system to disengage
when a quick stop is needed.
The more complex vehicles become,
with ever-more interconnected MCUs
needing to transfer information, the
more important the reliability of the
vehicle bus becomes. And with each
model year bringing new cameras,
sensors, and display screens, the
efficiencies that CAN provides in the
physical layer of a vehicle become
more attractive. In the past, cars were
limited in their features due to the finite
amount of space for the physical cables
and complex wiring that was required
for each system to communicate. CAN
allows for a leaner networked system
that not only underlies the connected
vehicles of today, but also the drive-
by-wire functionality necessary for the
autonomous vehicles of tomorrow.
How is CAN Being Used in
Military and Defense Vehicles?
As previously mentioned, the CAN
bus has been the communication
standard for embedded systems
The CAN bus allows for communication between embedded systems within a UAV, as well as the
transfer of information between a UAV and the remote operator.
UNMANNED SYSTEMS SPECIAL REPORT
in vehicles for decades, and even
huge leaps in vehicle technology like
electric and autonomous vehicles
have continued to utilize the CAN bus
due to its flexibility and reliability.
These same features make CAN an
ideal component for autonomous
military and defense vehicles,
including UGVs and unmanned
aerial vehicles (UAVs), or drones.
Military Vehicles
Commercial autonomous vehicles
must have highly attuned sensors
when navigating city streets — able
to sense changing road conditions,
other vehicles, and pedestrians.
Tactical military vehicles, on the other
hand, must be prepared for off-road
conditions in every kind of hostile
environment. The obstacles are greater
and the consequences are higher
stakes. That means that a higher
priority must be placed on sensors
and algorithms that can calculate
and make split-second decisions;
and the need for near-instantaneous,
error-free communication is critical.
CAN enables all of these complex
systems to communicate with the
clarity and speed that are necessary
when lives are on the line.
In addition to its functionality,
CAN’s inherent ruggedness is a
clear draw, performing just as
consistently in extreme heat and cold
as it does in arid and dusty climates
and extremely wet conditions.
What Sensors Are Attached
to the CAN Bus?
The technologies that enable
autonomous driving can vary slightly,
but they all require advanced vision
and sensing equipment to “see” the
road ahead, as well as high-powered
software to make decisions based
on that visual information. Most
autonomous military vehicles would
support some combination of the
following sensors, among others:
• Light Detection and Range
(LiDAR) technology, for creating
a 3D map of the road ahead;
• Color cameras, for determining the
changing position of the road and
other obstacles in front of the vehicle;
• Infrared cameras, adding
another layer of complexity
to obstacle-sensing; and
• GPS, for navigation and creating
a larger contextual map that
the vehicle can reference.
With autonomous vehicles, and
especially autonomous tactical
vehicles, the in-vehicle networks
supporting the advanced vision and
sensing technologies require a higher
bandwidth connection like those
provided by Ethernet or FlexRay. But
these connections can combine with
CAN or CAN FD (CAN with flexible data
rate) to create a robust network that
is flexible when performing tasks that
require high data throughput, and that
is quick and reliable when performing
more simplified communication tasks.
Many military vehicles make use
of the CAN bus to log and transfer
periodic operational data that are
reviewed by maintenance personnel
(or, more likely, computer algorithms)
for predictive maintenance — in other
words, analyzing operational data to
look for potential vehicle maintenance
issues so that they can be addressed
before they become critical.
MARCH 2020 11
The CAN Bus
CAMERA
RADAR
INFRARED SENSORS
This illustration
THIS ILLUSTRATION
REPRESENTS
Drones
The use of the CAN bus isn’t
just limited to UGVs. Unmanned
aircraft systems (UASs) have also
adopted CAN technology for its
low-latency, reliable communication
capabilities. In fact, there’s even
a UAVCAN protocol designed for
aerospace and robotic applications.
The CAN bus allows for
communication between embedded
systems within a UAV, as well as the
transfer of information between a
UAV and the remote operator. For
instance, the CAN bus can allow
the flight controller to manage the
throttle on the electronic speed
controller (ESC), but it also allows
the ESC to return hard real-time data
to the flight controller, including
temperature, amperage, voltage,
warning signals, etc. via live telemetry.
The real-time data, transferred
within microseconds, allows
remote pilots to react immediately,
12 MARCH 2020
LIDAR GPS
DATA COLLECTION
AND PROCESSING
represents an
ANexample
EXAMPLElayout of sensors
LAYOUT on a military
OF SENSORS
ON A MILITARY
making for much safer and more
reliable UAV flight operations.
MilCAN
To account for the many issues
specific to military vehicles, a working
group of the International High Speed
Data Bus-Users Group (IHSDB-UG)
developed the MilCAN higher layer
protocol in 1999, with the goal of
creating a standard interface for
utilizing CAN bus in military vehicle
development. There are two versions
of MilCAN: MilCAN A and MilCAN B.
Widely used in armored vehicles,
MilCAN A uses 29-bit identifiers and uses
a similar frame format to SAE-J1939.
Mission-critical in mind, MilCAN A
prioritizes message transmission, and
defines 1-Mbit, 500-Kbps, and 250-Kbit
communication rates.
MilCAN B is actually an extension of
the CANopen application layer, using
11-bit identifiers and only periodically
allowing data to be transmitted via
ULTRASONIC
SENSORS
vehicle. VEHICLE.
the bus. MilCAN B supports data
rates from 10 kbps to 1 Mbps.
Both protocols were developed
to specialize the use of CAN around
deterministic data transfer, so the
specifications can also be used
for non-military applications.
The Future of CAN in
Autonomous Military Vehicles
Don’t expect robotic armies to
be taking over the front lines any
time soon, but with successful pilot
programs like CAAR demonstrating the
viability of heavy-duty military grade
autonomous vehicles, there is hope for
making extremely risky situations — like
resupply missions — a little safer. And
for the foreseeable future, the CAN
bus will be along for that journey.
This article was written by
Jesse Paliotto, director of marketing,
Kvaser (Mission Viejo, CA).
For more information, visit
http://info.hotims.com/72993-501.
UNMANNED SYSTEMS SPECIAL REPORT
 Launch to new heights, achieve global recognition and take off with $20,000. You can
make your dream a reality by entering our 2020 Create the Future Design Contest.
By submitting your most innovative product ideas, you could shape the future!

Visit: CreateTheFutureContest.com
Entries: Accepted beginning March 1, 2020
See the website for contest rules and regulations

PRINCIPAL SPONSORS CATEGORY SPONSOR SUPPORTING

SPONSOR

PRIZE SPONSOR
AUTONOMOUS
“WINGMAN” By 2025, the Army sees ground troops conducting
foot patrols in urban terrain with robots, called Squad

VEHICLES
Multipurpose Equipment Transport vehicles, that carry
rucksacks and other equipment alongside soldiers.
Overhead, unmanned aircraft will also serve as spotters
to warn troops so they can engage the enemy on their
own terms, according to the Army’s new strategy on

The Future of Military Unmanned

robotic and autonomous systems.

Vehicle Technology

T
he US Army’s Futures The questions are many: What is The article examines the type of
Command is the most important to be the deployment doctrine and sensors, control, and telemetry needed
administrative reorganization of the sensors capabilities? Are these between an autonomous vehicle and

BACKGROUND PHOTO ILLUSTRATION BY PEGGY FRIERSON;

the modern Army. Responding
to the world’s changing priorities —
especially the “near peer” threat of
ascendant Russia and China — the
Army is no longer modernizing, but
re-inventing its ground vehicle fleet
against new realities. Just like the U.S.
Air Force stopped inventing better jets
and pilot aids and moved to unmanned
aerial vehicles (UAV) for “dull, dirty and
dangerous” missions, the Army envisions
multiple autonomous vehicle concepts.
new vehicles armed or merely for
C4ISR? And indeed, are they actually
“autonomous” or controlled, like
USAF UAVs, by remote operators
from mobile ground stations? The
good news is, regardless of how the
doctrine or vehicle takes shape, the
Army’s own MVD (Multi-function Video
Display) system can be “appliqued”
from the Type II MRAP Mine Clearing
Vehicle and used to remotely
control all aspects of tomorrow’s
its “chase Mobile Ground Station,” and
describes the capabilities of the MVD
on MRAP vehicles and how it directly
applies to the battlefield of the future.
Brief references to Next Generation
Combat Vehicles will be made, although
at time of writing, this program is
still evolving and far from settled.
National Defense Strategy Stays
Ahead of Russia and China
In less than 20 years, the US has
INSET: U.S. ARMY

Instead of a heavier Abrams main battle
tank, or a replacement to the aging M113
APC, autonomous “wingman” vehicles
may replace some of the human-
heavy tasks on the future battlefield.
“wingman” autonomous vehicles.
In this article, we briefly examine
what the Army’s Futures Command
has published on future vehicle fleet,
autonomy, and the evolving use cases.
changed its defense doctrine multiple
times in response to world events like
9/11, urban warfare in Iraq, fighting
ISIS — and now, a realization that
Russia and China are legitimate

14 MARCH 2020 UNMANNED SYSTEMS SPECIAL REPORT

 battlefield foes. While the first part of
the Department of Defense’s mission
statement says America still intends
to effectively fight in two battle
theaters simultaneously (a carryover
from World War II), the second half of
the mission statement guaranteeing
to “prevail using overwhelming
force” has DoD planners worried.
In early 2018, the Pentagon’s National
Defense Strategy (NDS) recognized
that on land, sea, and in the air, the
reality is that both Russia and China
pose serious threats. Military planners
believe there is no longer 100 percent
confidence in America’s technical and
weapons superiority. In Syria, DoD
planners watched Russia “test bed”
new offensive and defensive strategies
and technology and concluded
that America needed to re-evaluate
many of our deployed platforms.
The Army brass knew something had
to change. While weapons like the M1A2
Abrams main battle tank were designed
for head-to-head European combat
against Russian tanks, the reality shown
in Crimea is that Russian EW and cyber
capabilities will affect Army infrastructure
sooner than Russian armored vehicles.
Apparently with little effort, Russia was
able to “see” and “kill” the opposing
forces’ command and control structure.
Currently using a fixed and slow-to-
erect Forward Operating Base (FOB)
concept dating to World War I/II, the
Army has realized that speed — coupled
with a “shoot and scoot” model — is
now essential for modern warfare.
New Army Cross-Functional
Teams (CFTs) and NGCV
In late 2017, Army Chief of Staff
Mark Milley outlined a set of eight
cross-functional teams (CFTs)
designed to address new battlefield
realities, and more importantly, to
bring essential capabilities right to
the warfighter as quickly as possible.
The CFTs are listed in Figure 1.
The CFTs bring together multiple
ZWICKROELL
Army organizations, labs, vendors, and
industry to openly discuss out-of-the-
box ways to address the (primarily)
Russian threat. From a ground
UNMANNED SYSTEMS SPECIAL REPORT
vehicle perspective, platforms will
have increased lethality and mobility,
effectively inter-communicate in a
denied network and GPS environment,
and bring back the “overwhelming
force” part of the DoD’s mission
statement. In the past, bigger was
better. In the near future — by 2025,
in fact — smaller, more nimble,
networked, and autonomous vehicles
will enter the ground vehicle fleet.
Sensors will be an essential enabler to
realize an autonomous vehicle fleet.
At the Future Ground Combat
Vehicles Summit near TACOM in
Detroit, Michigan, in December
2018, Col. Warren Sponsler, Deputy
Director, NGCV CFT, Army Futures
Command, contrasted the difference
between today’s one-on-one/
vehicle-to-vehicle approach versus
the future multi-domain approach
(Figure 2, upper left frame).
In the lower frame, each ground
vehicle is networked to all other joint
battlefield assets. This is essentially
no different from today’s digital
battlefield concept, except it’s still
not a reality. Joint assets are rarely
directly inter-connected due to
interoperability constraints, relying
instead on intermediaries like FOBs,
AWACS, E2C, and other human
latency-prone “switch” points. In
the upper right frame, however, the
Army Cross Functional Team
Long-Range Precision Fires (LRPF)
Next Generation Combat Vehicle (NGCV)
Future Vertical Lift (FVL)
Network (including Precision Navigation
and Timing PNT)
Air and Missile Defense (AMD)
Soldier Lethality (including Synthetic
Training Environment STE)
Precision Navigation & Timing
Synthetic Training Environment
Figure 1. The U.S. Army’s Cross Functional
Teams (CFTs). Note that the Next Generation
Combat Vehicle isn’t a single vehicle, but a set
of at least five different platform types.
NGCV CFT envisions new platforms
supplementing ground vehicles by
providing reconnaissance and offensive
strike. Shown here, autonomous drone
swarms complement the indicated
autonomous vehicles. Both are inter-
linked to human-operated ground
vehicles and dismounted soldiers.
Enter NGCV: Not One, but Five
While NGCV is one of the Army’s
top eight CFTs, NGCV itself is not a
single vehicle but instead a set of CFTs
that is creating over five different new
ground vehicle concepts that all rely on
sensors like cameras, LiDAR, multiple
battlefield networks, and massive on-
board processing and digital storage.
Designed to supplement or replace
some current weapons platforms like
the M113 and Bradley Fighting Vehicle
(BFV or M2), the concept of Manned
Unmanned Teaming (MUM-T) currently
used on the Army’s Apache helicopter
will become a ground reality as long as
the technology enables the platform.
Each of the five NGCV CFT concepts
includes some form of autonomy.
The Robotic Combat Vehicles (RCV)
shown in Figure 3 are light, medium,
and heavy RCVs designed to act as a
“wingman” to other ground vehicles.
Used ahead of a strike force, the RCVs
might enable reconnaissance, offensive
weaponry, mobile FOB/command
post capability, and even “feint”
operations to draw an enemy away
from the main Army fighting force.
Technology Required
for Autonomy
Despite the positive press in the
commercial world about self-driving cars
from Google, Uber, Tesla, and others,
the DoD believes it will be the first to
actually deploy self-driving vehicles.
Michael Griffin, defense undersecretary
for research and engineering, told
Congress in April 2018: “We’re going
to have self-driving vehicles in theater
for the Army before we’ll have self-
driving cars on the streets.” To realize
his vision — plus the NDS top-down
goals, the Army’s top eight, and the
NGCV’s CFT's five vehicle-type portfolio
MARCH 2020 15
AUTONOMOUS
“WINGMAN” VEHICLES
— dramatic improvements in embedded,
deployed technology are needed.
That’s because the Army’s use case
is different from the commercial one.
Civilian self-driving cars require both a
Figure 2. The Army’s NGCV CFT’s vision for future warfare relies both on joint assets and Army-
owned autonomous platforms like UAS drones and autonomous ground vehicles.
Figure 3. NGCV portfolio consists of over five new ground vehicle platforms. The RCV concept relies
on manned/unmanned teaming (MUM-T) like the Army’s Apache strike helicopter.
16 MARCH 2020
sensor infrastructure built into roadways
and intersections, along with pre-defined
inter-vehicle protocols that allow vehicles
to talk to each other. On the battlefield,
there are no pre-defined roadways with
embedded sensors, and always-on inter-
vehicle communication would be either a
homing beacon for an enemy missile or
another network to be hacked or denied.
Instead, onboard sensors in Robotic
Combat Vehicles (RCVs) need to see,
understand, and navigate the terrain
around them. High-resolution closed-
circuit TV (ccTV) cameras mounted on
the vehicle will provide a 360-degree
view and embedded computers will
fuse the scene into a digital “cylinder”
as if an operator was looking all around
him inside the cylindrical scene.
In some cases where the RCV is in a
leader-follower role, a remote operator
might actually be viewing the scene
in real time, but onboard the RCV,
machine learning is compressing the
video while reducing it to predictable
quadrants to save RF bandwidth. In
this role, passive sensors like ccTV,
LiDAR, GPS and position/navigation
(pos/nav) using EMP-resistant
fiber optic gyros will be used.
Artificial intelligence (AI) — different
from machine learning by predicting
outcomes never previously encountered
or reasonably linearly extrapolated —
will go many steps further by merging
myriad battlefield sensors, metadata,
onboard databases, and occasional
use of active sonar and radar sensors.
By consuming as much data from
any available battlefield sensor feed,
whether on the vehicle or not, AI
NGCV CFT AND FUTURE GROUND COMBAT VEHICLES SUMMIT 2018
can autonomously drive the vehicle
without any operator input beyond a
designated destination or objective.
The more sensors and data added to
the vehicle, the embedded computer
challenge grows exponentially. For
example, while a quad core, 8th
Generation Intel Core i7 (Coffee Lake)
small form factor computer can decode/
encode streaming HD H.254 video in real
time at 60 Hz, it will struggle with two
simultaneous HD streams. If each vehicle
requires two HD cameras per corner
(90-degree field-of-view per camera,
with overlap), these eight HD video feeds
would require server-class processing
just for the CODECs. Image processing
requires even more processing power,
lower latency, and faster networks.
UNMANNED SYSTEMS SPECIAL REPORT
 Image processing — sensor fusion,
enhancement, anti-aliasing, target
tracking, compression, and identifying
areas of interest — now becomes a
digital signal processing challenge
requiring a dozen or more Intel
Xeon-class server cores and multiple
high-performance computing (HPC)
chassis, all small enough to be mounted
inside the vehicle. Onboard networks
from cameras and sensors to the
servers and operator displays need
at least gigabit throughput and likely
10 Gbits/s for sensor growth. Finally,
since these are forward-deployed
RCVs, they will likely sustain battle
damage, so fault-tolerant processors
and redundant in-vehicle networks
are needed — requiring multi-gigabit
Ethernet switches and mass storage.
MVD: Right, and Right Now
Figure 4 shows the DoD’s “Unmanned
Systems Integrated Roadmap: 2017-
2042.” The latest version, it identifies
numerous key technologies needed for
future autonomous vehicles. (Table 1)
In 2017, General Micro Systems
announced a contract win with the Army
for the Multi-Function Video Display
system (MVD) designed to bring video
and sensor data to Type II Mine Clearing
Vehicles on the MRAP chassis. Accepting
input from multiple high-resolution video
cameras, FLIR and the Vehicle Optic
Sensor System (VPOSS) — all called
“vehicle enablers” — the system converts
real-time data to ultra-low-latency
video-over-IP and transmits it over an
industry-standard 1 Gbits/s Ethernet
network, first to a rugged, embedded
Intel Xeon-class server and mass storage
system, then outward to multiple
operator smart screen consoles, each
with embedded Intel Core i7 processors.
Except for the sensors, GMS builds
100 percent of this intelligent, machine-
learning, open-standards network/
server/display system. The software
that brings the system to life is from the
Army’s Night Vision Labs and RDECOM
and is used to provide situational
awareness, operator hand off, and a
common user interface (UI) to disparate
operator stations and workloads.
UNMANNED SYSTEMS SPECIAL REPORT
MVD makes available today, not in
the future, many of the key technologies
identified in Table 1 that are essential for
RCVs as part of the Army’s NGCV CFT.
The Type II MRAP on which the system
is installed is not robotic, but MVD helps
control the platform’s robotic arm.
However, due to the under-one-frame
latency of MVD, the vehicle’s operators
can drive the vehicle completely “head
down” by only looking at the display.
This is possible without inducing nausea
because the under-one-frame latency
presents outside views practically in
real time. This capability is absolutely
essential for NGCV leader-follower
vehicles where operators are looking out
of sensors on the RCV but are located
separately and distant from the RCV.
GMS is working on enhancements to
MVD that will tick off the list even more of
the technologies in Table 1. For example,
the GMS X422 “Lightning,” announced
last October, brings AI to ground
vehicles by adding two Nvidia Tesla V100
GPGPU “data center accelerators” in a
conduction-cooled vetronics chassis. An
upgraded embedded server and open
standard network switch raise the in-
vehicle speed to 10 Gbits/s, providing
more bandwidth for high speed and
myriad more sensors. However, except
for the addition of the AI chassis (about
the size of a shoebox), the system
remains exceptionally small. It’s designed
to fit into ground vehicles wherever
there’s room: under a seat, mounted
vertically on a bulkhead, or tucked
behind a piece of existing equipment.
Robotics advancements
Prioritized common/open architectures
Common data repositories
Autonomous modeling and simulation
advancements
Machine learning advancements
Artificial Intelligence advancements
SWaP/miniaturization advancements
Swarming capabilities
Augmented Reality
Virtual Reality
Table 1. Key technologies identified by the DoD for unmanned systems.
UNMANNED SYSTEMS INTEGRATED ROADMAP
2017-2042
Statement A. Approved for public relase: distribution unlimited.
Figure 4. The DoD’s “Unmanned Systems
Integrated Roadmap” in 2017 charts the way
for all service branches.
NGCV and Autonomous
Vehicle Enablers
As this article has outlined,
autonomous “wingman” vehicles are
poised to replace many of the human-
heavy tasks on the future battlefield.
As the Army works to integrate the
most advanced technologies into
MVD on MRAPs and other Next
Generation Combat Vehicles, the US
will be more prepared to respond to
the “near peer” threats we are seeing.
This article was written by
Chris A. Ciufo, CTO, General Micro
Systems (Rancho Cucamonga, CA).
For more infor­mation, visit
http://info.hotims.com/72993-500.
Sensor advancements
Collision avoidance
Leader-follower
GPS-denied solutions
Cyber resilience and robustness
Information assurance solutions
Increased network and spectrum capacity
Human-machine interface advancements
Autonomous data strategy adaptation
MARCH 2020 17
MICROTURBINE PROPULSION
FOR UAVS
S
eventy years ago, military
aviation moved from
reciprocating engines to
vastly more reliable turbo jets
and turboprops. Shortly after, the
commercial air transport industry
followed suit, enabling modern air
transport. Today, virtually all large
aircraft rely on turbine propulsion,
yet small aircraft, both manned and
unmanned, have not exploited the
advantages of turbines for propulsion.
While UAVs have become integral
for both commercial and military
aerial missions, the continued use of
reciprocating engines has limited the
UAV market’s ability to reach its true
growth potential. The industry has
been slow to innovate and develop
turbine propulsion systems for small
aircraft because it is far more difficult to
design and produce high-performance
small turbines than large ones.
18 MARCH 2020
Today’s engine market is riddled Designing a Solution
with inefficiencies and hazards that After spending five years working
jeopardize the safety and reliability to design and manufacture a world-
of UAVs. Current engines powering class gas turbine engine small enough
drone delivery are unreliable and to fit into Group 3 (max gross takeoff
run on highly volatile
and dangerous fuel,
minimizing the impact of
a drone fleet and raising
costs. A major percentage
of vehicle losses with their
payloads are attributed
to engine failure. Plus,
with the need for frequent
overhauls, customers
have to purchase multiple
engines for a single
vehicle so that, when one
engine is being worked
on, they can continue
operating with an
alternate engine. This severely limits Artist’s illustration of a Monarch RP
the potential for the entire industry. microturbine engine.
UNMANNED SYSTEMS SPECIAL REPORT
weight less than 1320 lbs) and Group
4 (MGTOW of more than 1320 lbs), the
team at UAV Turbines has achieved
this engineering feat with the launch
of its Monarch propulsion system.
The Monarch turboprop was carefully
designed to outperform conventional
reciprocating engines in several ways:
• Monarch propulsion systems will
provide more time in the air and less
time being serviced on the ground with
upwards of a 2,000 hour increase in
operation time between overhauls when
compared to available Class 3 engines.
• The Monarch system’s variable
pitch propeller will enable UAVs
to climb faster and reach greater
dash speeds, enabling greater
performance and efficiency in both
commercial and military aircraft.
• The reliability of Monarch systems
eliminates the need for extra
engines for a single aircraft.
• The flexibility to run efficiently on
all types of heavy fuels, such as jet
fuel, makes Monarch propulsion safer
and more convenient than engines
running on volatile aviation gasoline.
• Monarch RP generates useful on-
board electrical power that is 2-3x
greater than what is produced
by conventional engines.
In designing this system, the team
at UAV Turbines had to overcome
two key issues facing the UAV engine
industry: designing a small, powerful
engine that was both reliable and
ran on safer fuel. There are several
fundamental problems that had to be
overcome, starting with the fact that
the size and weight of these systems
are major constraints. A little over
1,300 pounds really isn’t much.
How It Works
To achieve significant thrust, the heavy
fuel (typically JP-8 or Jet A) burns in a
very small space at similar temperatures
as the larger engines. The internal
distances, however, are much smaller,
so managing thermal gradients and the
resulting stress becomes more difficult.
The low air flows in these engines call
for the use of very small air passages
and very high-speed turbomachinery.
UNMANNED SYSTEMS SPECIAL REPORT
Impeller and turbine for a 197,000 RPM turbogenerator, one of several 10 HP electrical generator
prototypes built and operated in the UAVT development program.
The core rpm in the first system
UAV Turbines is releasing for flight
test is right around 100,000 rpm
at cruise speed. UAV Turbines built
even smaller experimental engines
with rotor speeds of 200,000
rpm. Managing tolerances and
designing to process capabilities
becomes critical. New engineering
approaches were essential to deal
with the extreme internal thermal
stresses and tight tolerances.
One solution to the thermal issue
was to separate the hot section
from the engine’s bearing cavity by
placing the turbine rotor at the end
of the shaft, cantilevered, so that
the bearings are located in a cold
section of the engine core. Overhung
systems have been used before but
there are critical fits between the
components that have to provide
intimate contact from zero cold to
full speed hot — and all transients
in between. The designers ended up
balancing on the limits of what can
be made with what will operate.
Another factor is that this is not just
an engine, but a propulsion system. To
put that power to work, it’s necessary
to step down from 100k core rpm
to 6k rpm via an extremely efficient,
lightweight gearbox. This drives a
variable pitch propeller so that the
power can be efficiently converted to
thrust through each mission stage, from
takeoff to landing. The engine itself
runs at essentially constant speed, so
pilot throttle changes result in pitch
changes to the propeller blades via a
variable pitch mechanism as calculated
by the engine FADEC (Full Authority
Digital Engine Control). The propeller
itself was designed with small pusher
UAV installations in mind. The propeller
was designed with three blades
to reduce noise from the propeller
interacting with the wing wakes,
increase ground clearance, and reduce
the chance of ground strikes during
takeoff or landing. The entire system
must then be integrated into a flyable
package, a classic systems challenge.
Today’s military and commercial
organizations are urgently in
need of a more reliable and safer
propulsion system for their UAVs.
Reliable, lightweight, fuel-efficient
microturbine engines may be the
answer to provide propulsion and
power generation in small to medium-
sized UAV propulsion systems
This article was written by Kirk
Warshaw, CEO, UAV Turbines (Miami,
FL). For more information, visit
http://info.hotims.com/72996-502.
MARCH 2020 19
EYES IN THE SKY
For Drones, Combining Vision Sensor and IMU
Data Leads to More Robust Pose Estimation
D
rones (i.e. quadrotors) are
a popular and increasingly
widespread product used
by consumers as well as in
a diversity of industrial, military, and
other applications. Historically under
the control of human operators on the
ground, they’re becoming increasingly
autonomous as the cameras built into
them find use not only for capturing
footage of the world around them
but also in understanding and
responding to their surroundings.
Combining imaging with other
sensing modalities can further bolster
the robustness of this autonomy.
When a drone flies, it needs to
know where it is in three-dimensional
space at all times, across all six
degrees of freedom, for translation
and rotation. Such pose estimation
is crucial for flying without crashes
or other errors. Drone developers
are heavily challenged when
20 MARCH 2020
attempting to use a single IMU
or vision sensor to measure both
orientation and translation in space.
A hybrid approach combining IMU
and vision data, conversely, improves
the precision of pose estimation for
drones based on the paired strengths
of both measuring methods.
The IMU sensor measures
acceleration, with information about
the orientation derived from its
raw output data. In theory, such
acceleration measurements could also
be used to derive translation. However,
to calculate such results, developers
need to integrate twice, a process that
results in increased errors. Therefore,
the IMU sensor alone is not an accurate
source of precise location information.
In contrast, the vision sensor is quite
good at measuring location; it’s sub-
optimal at determining orientation,
however. Particularly with wide-view
angles and long-distance observation,
it’s quite complicated for the vision
system alone to measure orientation
with adequate precision. A hybrid
system of paired IMU and vision
data can provide a more precise
ONE/SHUTTERSTOCK.COM
measurement for the full six degrees
of pose in space, providing better
results than using either the IMU
or the vision sensor individually.
The most challenging issues in
such a sensor fusion configuration
are to determine a common
UNMANNED SYSTEMS SPECIAL REPORT
 coordinate frame of reference for
both orientation and translation
data, as well as to minimize the noise
produced by the sensors. A common
approach to create the reference
frame leverages linear Kalman
filters, which have the capability to
merge both IMU and vision data for
hybrid pose estimation purposes.
For a vision system mounted on
or embedded in the drone, SLAM
(simultaneous localization and
mapping) provides spatial awareness
for the drone by mapping its
environment to ensure that it does
not collide with trees, buildings,
other drones, or other objects.
Factors to Consider When
Building a Hybrid Sensor-Based
Drone System
Several key factors influence
the measurement quality. First
off, the quality results of an IMU’s
measurements highly depend on
the quality of the IMU selected
for the design. Inexpensive IMUs
tend to generate high noise levels,
which can lead to various errors and
other deviations. More generally,
proper calibration is necessary to
comprehend the filter’s characteristics,
e.g. the sensor’s noise model.
Individual sensors, even from the
same model and manufacturer,
will have slightly different noise
patterns that require consideration.
On the vision side, the
implementation specifics fundamentally
depend on whether a global or rolling
shutter image sensor is being used.
With a global shutter image sensor,
every pixel is illuminated at the same
time, with no consequent readout
distortion caused by object motion.
With a more economical rolling
shutter image sensor, conversely,
GEOGIF/SHUTTERSTOCK.COM
distortion can occur due to read-
out time differences between pixels.
IMU information can correct for
rolling shutter artifacts, historically
by the use of various filter methods.
Nowadays, the noise reduction of the
IMU sensor can also be corrected by
deep learning-based processing.
UNMANNED SYSTEMS SPECIAL REPORT
Combining IMU
and Vision Data
One challenge
with hybrid systems
is that the captured
vision data is
often very slow
from a frame rate
standpoint, usually
well below 100 Hz,
while the IMU data
comes across at
high frequency,
sometimes well
over 1 KHz. The root Considering the limited integration space in most drones,
of the resultant implementing a robust setup for synchronization of the IMU and
implementation vision data can get complicated.
problem lies in
finding a way to obtain information
from both systems at the exact
same time. SLAM techniques such
as Continuous Trajectory Estimation
can approximate the drone’s
movement by assuming that the
drone’s speed is continuous.
Developers can integrate both
the IMU and vision data into an
environment with a common
reference frame, allowing them to
assign measurements to a specific
part of the continuous trajectory. In-
between any two image acquisitions,
multiple IMU measurements provide
additional reference points regarding
this trajectory. When in the air,
the drone will then constantly be
time-synchronized and updated
with IMU data. And every time a
vision image is received, it then
corrects the IMU information.
Hardware Requirements,
Implementation, and Testing
Considering the drones’ limited
integration space and more general
resource-limited embedded nature,
implementing a robust setup for syn-
chronization of the IMU and vision data
is not straightforward. Light and lean
components with powerful processing
units are necessary in a limited memory
footprint while consuming little power.
Stabilized sensors and software filtering
are also essential for SoC developers
as a result of the extreme jitter due to
propeller movement, which influences
the vision components of the design.
Both the vision sensor and IMU have
individual local reference systems to
measure their own pose, which need
to be calibrated to ensure a hybrid
SLAM has a common reference frame.
Each system must be calibrated
individually first, and then co-calibrated
with respect to each other, in order to
receive the position in the common
reference frame. Multiple data sets are
available to test aerial vehicle pose
estimation, also in hybrid fashion,
with the developed pipelines. The
most commonly used dataset is
EuRoC, which provides raw vision
and IMU data to test algorithms and
compare against other methods.
While visible light image sensors
are often included in autonomous
system designs, they’re not necessarily
the sole required sensor technology
in all implementation situations. By
combining them with other sensor
technologies, however, the resultant
“sensor fusion” configuration can
deliver a robust implementation in all
possible usage scenarios, such as for
semi- and fully autonomous vehicles,
industrial robots, drones, and other
autonomous device applications.
This article was written by Ute
Häußler, corporate editor, FRAMOS
GmbH (Taufkirchen, Germany).
For more information, visit
http://info.hotims.com/72993-504.
MARCH 2020 21
APPLICATION BRIEFS
Parcel Delivery Drone
Airbus Helicopters
Marignane Cedex, France
+33 (0)4 42 85 60 51
http://helicopters.airbus.com

A irbus has begun shore-to-ship trials in Singapore

with its Skyways parcel delivery drone. This marks
the first time drone technology has been deployed in real
port conditions, to deliver a variety of small, time-critical
maritime essentials to working vessels at anchorage.
The maiden shore-to-ship delivery flight was made
to the Swire Pacific Offshore’s Anchor Handling Tug
Supply vessel ‘M/V Pacific Centurion’, 1.5km from the
shoreline of Singapore’s Marina South Pier, carrying
1.5kg of 3D-printed consumables. Landing safely on the
ship deck and depositing its cargo to the shipmaster,
the Skyways unmanned air vehicle swiftly returned to its reducing carbon footprint, and significantly mitigating
base, with the entire flight taking less than ten minutes. risks of accidents associated with launch-boat deliveries.
The trials are being undertaken in conjunction with Airbus and Wilhelmsen Ships Services signed an agreement
partner Wilhelmsen Ships Services, one of the world’s in June 2018 to drive the development of an end-to-end
leading maritime logistics and port services companies. unmanned aircraft system for safe shore-to-ship deliveries.
During the trials, Airbus’ Skyways drone will lift off from The collaboration marries Airbus’ expertise in aeronautical
the pier with a payload capability of up to 4kg and vertical lift solutions and Wilhelmsen’s wealth of experience
navigate autonomously along pre-determined ‘aerial in ship agency services. A landing platform and control
corridors’ to vessels as far as 3km from the coast. center were set up at the Marina South Pier in November
The use of unmanned aircraft systems in the maritime 2018, through the facilitation of the Maritime and Port
industry paves the way for possible enlargement of existing Authority of Singapore. The maritime agency also designated
ship agency services’ portfolio, speeding up deliveries anchorages for vessels to anchor off the pier for the trials,
by up to six times, lowering delivery costs by up to 90%, while the Civil Aviation Authority of Singapore worked with
Airbus and Wilhelmsen to
ensure safety of the trials.
Skyways is an experimental
project aimed at establishing
seamless multi-modal
transportation networks in smart
cities. Through Skyways, Airbus
aims to develop an unmanned
airborne infrastructure solution
and address the sustainability
and efficiency of unmanned
aircraft in large urban and
maritime environments. Having
demonstrated the ability to
deliver parcels safely and reliably
to vessels anchored off the
coast of Singapore, Skyways will
soon be commencing another
trial phase delivering air parcels
autonomously in an urban
environment, at the National
University of Singapore.
For Free Info Visit http://
info.hotims.com/72993-462.

22 MARCH 2020 UNMANNED SYSTEMS SPECIAL REPORT

Robotic Helicopter
Steadicopter Ltd.
Migdal HaEmek, Israel
+972-4-9592959
www.steadicopter.com
S teadicopter, one of the companies involved in the
Rotary Unmanned Aerial Systems (RUAS) industry,
recently unveiled its next generation Black Eagle 50
advanced lightweight unmanned robotic helicopter. New
capabilities include an inertial navigation system with no
dependence on GPS, as well as support for naval missions.
Steadicopter’s Black Eagle 50 unmanned helicopter
has been upgraded with several additional new features
and is tailored for naval missions with its robust
mechanical and electronic capabilities that support flight
in maritime environmental conditions. The company
also recently signed a cooperative agreement with Israel
Shipyards for the marketing of the Black Eagle as part
of the defense, intelligence, and surveillance systems
installed on its OPV (Offshore Patrol Vessel) family. The
OPV vessels, based on the Nirit Class SAAR 4.5, are
heavily armed with advanced weapons systems and
can be equipped with helicopter carrying capability.
UNMANNED SYSTEMS SPECIAL REPORT
The Black Eagle 50 features a special inertial navigation
system capability, based on input from the system’s inertial
and other sensors. Through a unique smart navigation
algorithm, this input enables the continuation of the flight
and the mission without relying on GPS, giving the Black
Eagle 50 a significant advantage in GPS-denied areas.
In terms of mission capability, the Black Eagle 50 is a VTOL
(Vertical Take-Off and Landing) robotic observation system,
suitable for tactical maritime and land Intelligence, as well as
Surveillance, Target Acquisition and Reconnaissance (ISTAR)
missions. A data link connects the aircraft with the ground
controller, enabling the transmission of live video and data
between the two. The system has a steady hovering endurance
of up to 3 hours and flight endurance of up to 4 hours.
At only 27 kg, the Black Eagle 50 is extremely lightweight
and compact, with a maximum takeoff weight of 35 Kg, and
payload capacity of 5 Kg. It has a communication range
of up to 150 km, depending on the client’s requirements,
and a service ceiling of up to 10,000 ft. Its total length is
just 2540 mm, while its maximum air speed is 70 knots
(126 Km/h) with a cruising speed of 45 knots (81 Km/h).
For Free Info Visit http://info.hotims.com/73000-464.
MARCH 2020 23
TECH BRIEFS
Aircraft Vertical Takeoff and Landing
A vertical-takeoff UAV enables long-endurance missions and is easy to transport.
Langley Research Center, Hampton, Virginia
N ASA’s Langley Research
Center de­veloped
an inexpensive, long-en­
durance, vertical takeoff and
landing (VTOL) unmanned
aerial vehicle (UAV).
It is capable of flying for 24
hours, landing in a 50 × 50
zone, and can be loaded into
the back of a cargo van for
easy transport. In addition, it
can land in either a horizontal
or vertical flight configuration.
The vehicle can be used
in any application where The unique design of the aircraft enables VTOL capabilities.
UAVs are needed to take off
from, or land in, an unprepared landing
area. Commercial applications where
extended-mission UAV capabilities are
beneficial include law enforcement,
firefighting, crop surveys, pipeline
surveys, or oil field management.
Neural Lander Uses AI to Land Drones Smoothly
The system employs a deep neural network to overcome the challenge of ground-effect turbulence.
California Institute of Technology, Pasadena
L anding multi-rotor drones smoothly
is difficult. Complex turbulence
is cre­ated by the airflow from each
rotor bouncing off the ground during
a descent. This turbulence is not well
understood nor is it easy to compensate
for, particularly for autonomous drones.
That is why takeoff and landing are
often the two most difficult parts of a
drone flight. Drones typically wobble
and inch slowly toward a landing until
power is finally cut and they drop the
remaining distance to the ground.
A system was developed that uses
a deep neural network to help
autono­mous drones “learn” how to
land more safely and quickly while
using less power. The system, called
the Neural Lander, is a learning-
based controller that tracks the
position and speed of the drone and
24 MARCH 2020
This VTOL aircraft is a modification
of a conventional single-prop aircraft
design. The addition of vertically
oriented, stowable tail rotors and an
articulating forward rotor capable of
pivoting from a horizontal to vertical
modifies its landing trajectory and
rotor speed to achieve the smoothest
possible landing. It has the potential
to help drones fly more smoothly
and safely, espe­cially in the presence
of unpredictable wind gusts.
Deep neural networks (DNNs) are
AI systems inspired by biological
systems like the brain. The “deep”
part of the name refers to the fact
that data inputs are churned through
multiple layers, each of which
processes incoming information
in a different way to tease out
increasingly complex details.
DNNs are capable of automatic
learning, which makes them ideally
suited for repetitive tasks.
To make sure that the drone flies
smoothly under the guidance of the
DNN, a technique known as spectral
orientation enables VTOL
capabilities. It combines the
speed and fuel efficiency
of fixed-wing aircraft
with the hoverability and
flexibility of rotary aircraft.
The design is expected to
avoid the compromises
in performance that
are typically made
in development of
VTOL aircraft.
NASA is actively
seeking licensees to
commercialize this
technology. Please contact
The Technology Gateway at LARC-
DL-technologygateway@mail.nasa.
gov to initiate licensing discussions.
Follow this link for more information:
https://technology.nasa.gov/
patent/LAR-TOPS-230.
normalization was used that smooths
out the neural net’s outputs so it doesn’t
make wildly varying predictions as in­
puts or conditions shift. Improvements
in landing were measured by examining
deviation from an ideal trajectory in
3D space. Three types of tests were
conducted: a straight vertical landing,
a descending arc landing, and flight in
which the drone skims across a broken
surface, such as over the edge of a
table, where the effect of turbulence
from the ground would vary sharply.
NASA/WILLIAM J. FREDERICKS
The new system decreases vertical
error by 100 percent, allowing for
controlled landings, and reduces
lateral drift by up to 90 percent. In
experiments, the system achieved
actual landing rather than getting
stuck about 10 to 15 centimeters
above the ground, as unmodified
UNMANNED SYSTEMS SPECIAL REPORT
conventional flight controllers often
do. Further, during the skimming
test, the Neural Lander produced
a much smoother transition as the
drone transitioned from skimming
Lightweight Sensing and Control System for Unmanned
Aerial Vehicle Monitoring
The control system allows semi-autonomous flight.
Langley Research Center, Hampton, Virginia
A new sensing and control
system for unmanned aerial
vehicles (UAVs) allows for semi-
autonomous flight. Pilots need not
leave the ground to conduct routine
monitoring and surveillance quickly
and cost-effectively. Such systems are
particularly useful during long flight
segments or over remote locations,
or for scientific applications such
as atmospheric monitoring or crop
monitoring, which might require long
and repeated sampling in a specific
pattern. The small, lightweight
technology can be quickly adapted
to a specific configuration.
Increasing demand for smaller UAVs
(e.g., sometimes with wingspans on
Flying Robot Mimics Rapid Insect Flight
The autonomous, free-flying robot enables new drone applications.
Delft University of Technology, Delft, The Netherlands
F lying animals both power and
control flight by flapping their
wings. This enables small natural
flyers such as insects to hover close
to a flower but also to rapidly escape
danger. Animal flight has always
drawn the attention of biologists, who
not only study their complex wing
motion patterns and aerodynamics
but also their sensory and neuro-
motor systems during such agile
maneuvers. Recently, flying animals
have also become a source of
inspiration for robotics re­searchers,
who try to develop lightweight flying
UNMANNED SYSTEMS SPECIAL REPORT
across the table to flying in the
free space beyond the edge.
Besides its obvious commercial
applications, the new system could prove
crucial to autonomous medical transports
the order of 6" and weighing less than
one pound) generated a need for much
smaller flight and sensing equipment.
NASA Langley’s new sensing and flight
control system for small UAVs includes
both an active flight control board and
an avionics sensor board. Together,
these compare the status of the UAV’s
position, heading, and orientation with
the pre-programmed data to determine
and apply the flight control inputs
needed to maintain the desired course.
To satisfy the small form-factor system
requirements, microelectromechanical
systems (MEMS) are used to realize the
various flight control sensing devices.
MEMS-based devices are commercially
available single-chip devices that lend
robots that are agile, power-efficient,
and even scalable to insect sizes.
Researchers have developed the Del­
Fly Nimble, a novel insect-inspired flying
robot with flapping wings, beating 17 times
per second, that not only generate the
lift force needed to stay airborne but also
control the flight via minor adjustments
in the wing motion. Inspired by fruit flies,
the robot’s control mechanisms have
proved to be highly effective, allowing
it not only to hover on the spot and fly
in any direction but also be very agile.
The robot has a top speed of 25 km/h
and can perform aggressive maneuvers
that could land in difficult-to-reach
locations (such as gridlocked traffic).
For more information, contact
Robert Perkins at rperkins@caltech.edu;
626-395-1862.
themselves to easy integration onto
a circuit board. The system uses less
energy than current systems, allowing
solar panels planted on the vehicle to
generate the system’s power. While the
lightweight technology was designed
for smaller UAVs, the sensors could be
distributed throughout larger UAVs,
de­pending on the application.
NASA is actively seeking licensees to
commercialize this technology. Please
contact The Technology Gateway
at LARC-DL-technologygateway@
mail.nasa.gov to initiate licensing
discussions. Follow this link for more
information: https://technology.
nasa.gov/patent/LAR-TOPS-33.
such as 360-degree flips, resembling
loops and barrel rolls. The 33-cm
wingspan and 29-gram robot has, for its
size, excellent power efficiency, allowing
five minutes of hovering flight, or a
flight range of more than 1 km, on a fully
charged battery. The robot has a thrust-
to-weight ratio of more than 1.3 and is
capable of carrying an additional payload
of up to 4 grams (e.g. a camera system
with a live video feed, additional sensors,
etc.) The exceptional agility can be
demonstrated by 360-degree flips around
the pitch or roll axes or rapid transitions
from hover to forward or sideways flight,
MARCH 2020 25
TECH BRIEFS
The DelFly Nimble in stationary (hovering) flight.
and vice versa. At full throttle, the robot
reaches a top speed of 7 m/s (~25 km/h).
Like in quadrotors or helicopters, but
also like in insects, forward/backward
and sideways flight is achieved by
pitching and rolling the robot’s body
into the respective direction. To control
the body orientation (attitude), the
robot needs to be able to produce
torques around the three orthogonal
body axes. To this end, the robot is
equipped with two independent flapping
mechanisms — one for each wing pair
on the sides of the robot. These are
complemented with two rotary servo
actuators — one adjusting the dihedral
Software Enables Robots to Self-Right After Overturning
These robots feature an open systems architecture that enables them to get up after falling.
Army Research Laboratory, Adelphi, Maryland; Johns Hopkins University Applied Physics Laboratory, Baltimore, Maryland;
and Northrop Grumman Corp., Falls Church, Virginia
T he Advanced Explosive Ordnance
Dis­posal Robotic System (AEODRS)
features a modular open systems
architecture that enables the robot
to be self-righting after a fall. The
self-righting ability is ex­tended
to robots with a greater number
of degrees of freedom (joints).
The software looks at all possible
geometries and orientations that
the robot could find itself in. Each
additional joint adds a dimension
to the search space. This analysis
was made possible by the Range
Ad­versarial Planning Tool (RAPT),
a software framework for testing
autonomous and robotic systems.
26 MARCH 2020
Left wing-pair
Right wing-pair flapping mechanism
flapping mechanism
Dihedral angle
control mechanism
Autopilot &
radio receiver
Battery
Wing root
adjustment mechanism
The robot is equipped with two independent flapping mechanisms — one for each wing pair on the
sides of the robot. These are complemented with two rotary servo actuators — one adjusting the
dihedral angle by changing the relative orientation of the two flapping mechanisms and the other
actuating the tips of the left and right wing-pair roots.
angle by changing the relative orientation The robot has potential for novel
of the two flapping mechanisms and applications as it is lightweight, safe
the other actuating the tips of the left around humans, and able to fly more
and right wing-pair roots. Rolling is efficiently than more traditional drone
achieved by driving the two wing pairs designs, especially at smaller scales.
at different flapping frequencies, which For more information, contact Matěj
results in a thrust difference creating Karásek at M.Karasek@tudelft.nl;
the torque around the roll axis. +31 6 8384 1669.
An adaptive sampling algorithm
looks for transitions — states in which
the robot could transition from a
stable configuration to an unstable BELOW: NORTHROP GRUMMAN CORPORATION
one, thus causing the robot to tip
over. The software effectively predicts
where those transitions might be so
the space can be searched efficiently.
ABOVE: HENRI WERIJ, TU DELFT;
The AEODRS’s eight degrees of
freedom were analyzed and it was
determined that it can right itself on
level ground no matter what initial
state it finds itself in. The analysis
also generates motion plans showing
The Advanced Explosive Ordnance Disposal how the robot can reorient itself.
Robotic System Increment 1 Platform was used For more information, contact the ARL
to demonstrate the self-righting ability. Public Affairs Office at 301-394-3590.
UNMANNED SYSTEMS SPECIAL REPORT