Ground_Support_October_2020 - 27

GROUND SERVICE
PROVIDERS

"If you know the 3D geometry of vehicles
and aircraft and you can accurately identify
at least one physical point on each of these
objects plus their orientation, you have everything you need. With software that is similar
to that used for CAD, you can predict potential collisions or interferences of things before
they happen," Tabor says.
"If you were about to drive a baggage vehicle under a wing but you just happen to have a
conveyor up too high, the system could alert
you that, if you continue in that vector, you
will collide with the plane wing in 3D space,"
he adds.
The system would work much like a
backup sensor on a car, but without the need
for a camera to be on the GSE. Tabor explains
maps could be created inexpensively using
navigation technology used by drones and
then relayed to the GSE.
"Basically, it's a virtual 3D copy of the real
objects in real time. So as you're backing up,
the system knows that a part of your vehicle is
10 feet from a portion of a plane and a display
in your vehicle can show you an animated
image of what's behind you without actually
mounting cameras or other sensors that are
expensive and provide limited feedback," says
Tabor.
This technology could be further enhanced
by utilizing artificial intelligence (AI). Tabor
points to research done where AI was able
to predict that someone was about to come
around a door before they were within a camera's view by noticing minute details.
"It sounds impossible, but really it is that
the pixels just show a slight difference of
when a person gets close to being within view.
Their presence would produce ultra-subtle
reflections off the wall and the floor, stuff
your eyes would never even notice, the camera
system can detect," he says.
GSE could utilize AI to pick up, for example, the moment the shadow of a plane or person not in its original map came into view and
stop itself before an accident. Then through
noticing patterns and learning its environment, the GSE's AI could begin to predict
events and take action before they register to
anyone else on the ramp.
The AI would learn to take these steps over
constantly monitoring the actions taken and
learning the normal patterns of that GSE's
operations. Then when an action happened

that deviated from that norm, the AI would
step in to alert an operator or take steps itself
to prevent a potential accident.
"If an event happens that seems questionable and could cause a problem once, you can
record the event in the AI system as being
something not to do," Tabor explains. "When
things are always going right, it just keeps
on recording it in AI and then as soon as it
finds an anomaly, it gives you an alert that
you're doing something different than what
the normal is."
Tabor says they have designed a similar AI
system for a customer that performs pneumatics failure analysis and algorithms that
would predict failures.
"By using AI, you can train what is normal
and what is not," says Tabor.
Tabor says the technology to make it all
happen already exists. The only thing that
hasn't been done yet is to utilize it in the airport environment.

"All the technology that we need to accomplish this already exists, and so it's not like
having to develop a new class of expensive
sensors or anything else," Tabor says. "If you
can't obtain the 3D model from the manufacturer of the vehicle or plane, there is equipment available today that you can buy or rent
to create this 3D image yourself using LiDAR
or 3D cameras."
Tabor theorizes it would take less than six
months to scan everything and create a functional 3D map of a tarmac. The technology
could be easily retrofitted onto existing GSE.
"Most vehicles already have what is called
a CAN bus interface that makes available a lot
of that information such as speed, GPS location and heading. By interfacing that CAN
bus and fusing that data with other sensors
plus data from the 3D system, equipment like
a baggage handler vehicle, can implement this
technology without the need to add another
GPS or heading sensor," he says.

CONTAINER
DOLLY

SPECIFICATIONS
WE ARE HERE TO HELP!

WILCOX GROUND SERVICES
550 MCGEACHIE DRIVE
MILTON, ON L9T 3Y5 CANADA

WWW.WILCOXGROUNDSERVICES.COM

ACCOMMODATES
U LD 1, 2 & 3 MODELS
LOAD CAPACITY 3,500 LBS
STEEL CHANNEL FRAME
HOT DIPPED GALVANIZED BASE
POWDERCOAT TOP
PATENTED 4-WAY
STEERING DESIGN
EXCLUSIVE PRECISION
TRACKING DESIGN
SOLID RUBBER TIRES
TOWBAR ACTUATED
PARKING BRAKE

AviationPros.com/12390248

OCTOBER 2020

AviationPros.com

http://www.wilcoxgroundservices.com http://www.AviationPros.com/12390248 http://www.AviationPros.com

Ground_Support_October_2020

Table of Contents for the Digital Edition of Ground_Support_October_2020