Tech Briefs Magazine - August 2023 - 25

" Those computers are going to be
taking all of this new data we'll be acquiring
in real time through the 5G
system and running that through an AI
modeler to understand trends and
what the specific needs and requirements
will be for future refueling so
they can pre-plan and optimize operations, "
said Alvord.
The project is part of a larger initiative
within R&E to expand the use of 5G
technology throughout the Department
of Defense.
EA-18G Growlers on the ramp at Naval Air Station Whidbey Island. A pilot project under development
will create an optimized refueling system designed to boost readiness for military aircraft
operating there. (Image: U.S. Navy)
" Whidbey, through Joint Base Pearl Harbor-Hickam,
was directed to leverage commercially
available 5G networks, technologies,
and processes to experiment with
how to ensure that U.S. forces will have
connectivity uniquely suited to the battle
space wherever we deploy, " said Deb Stanislawski,
the OUSD (R&E) 5G Accelerate
Use Director. " Utilizing AI to assist in optimizing
5G network management to support
fueling operations is one of our experiments
that utilizes relevant mission
use cases that potentially can support real
world functionality and military utility. "
The new system will provide data to
three key groups: Fueling technicians
will know which aircraft need service and
what their priorities are; plane captains
will know when fuel trucks will be available
to service their aircraft so they can
be present, without having to wait on the
flight line; and base leadership will know
that available resources are being used
to keep missions on time.
" What 5G brings to this is decreased
latency - the ability to get more data in
real time - and increased bandwidth,
which allows us to get all the data we
need, " Alvord said.
The goal is for the new refueling system
to be fully implemented at Whidbey
no later than 2024 and to expand the
system to other sites.
" We're off to the races and getting
great feedback, " Alvord said. " There's a
desire to expand not just to other fueling
locations, but to apply what we're
doing to similar types of operations elsewhere
in the DoD, where having quick
access to large amounts of data and analytics
will be useful. "
For more information, contact Michelle
Gowdy at Michelle.Gowdy@gtri.gatech.
edu; 404-407-8060.
Drones Navigate Unseen Environments via Liquid
Neural Networks
A new level of autonomous drone navigation could enable real-world drone applications
like search and rescue, delivery, and wildlife monitoring.
Massachusetts Institute of Technology, Cambridge, MA
R
esearchers from MIT's Computer Science
and Artificial Intelligence Laboratory
(CSAIL) have introduced a method
for robust flight navigation agents to
master vision-based fly-to-target tasks in
intricate, unfamiliar environments. The
liquid neural networks, which can continuously
adapt to new data inputs,
showed prowess in making reliable decisions
in unknown domains like forests,
urban landscapes, and environments
with added noise and occlusion.
These adaptable models, which outperformed
many state-of-the-art counterparts
in navigation tasks, could enable
potential real-world drone applications
like search and rescue, delivery, and
wildlife monitoring.
Tech Briefs, August 2023
The researchers' recent study, published
in Science Robotics, details how this
new breed of agents can adapt to significant
distribution shifts, a long-standing
challenge in the field. The team's new
class of machine-learning algorithms,
however, captures the causal structure of
tasks from high-dimensional, unstructured
data, such as pixel inputs from a
drone-mounted camera. These networks
can then extract crucial aspects of a task
and ignore irrelevant features, allowing
acquired navigation skills to transfer targets
seamlessly to new environments.
" We are thrilled by the immense potential
of our learning-based control approach
for robots, as it lays the groundwork
for solving problems that
arise
www.techbriefs.com
when training in one environment and
deploying in a completely distinct environment
without additional training, "
said Daniela Rus, CSAIL Director and
the Andrew (1956) and Erna Viterbi Professor
of Electrical Engineering and
Computer Science at MIT.
" Our experiments demonstrate that
we can effectively teach a drone to locate
an object in a forest during summer, and
then deploy the model in winter, with
vastly different surroundings, or even in
urban settings, with varied tasks such as
seeking and following. This adaptability
is made possible by the causal underpinnings
of our solutions. These flexible algorithms
could one day aid in decision-making
based on data streams that change over
25

http://www.techbriefs.com

Tech Briefs Magazine - August 2023

Table of Contents for the Digital Edition of Tech Briefs Magazine - August 2023