What is DARPA looking for in the mini UAV, an “Unmanned Aerial Vehicle?”
Well, if you go to their website you’ll find (1) that DARPA knows what it
wants, and (2) it produces some long lists of “required” or “necessary” project
objectives.
But let’s take a step back and refocus on the big picture. DARPA is
looking for mini UAV’s with two characteristics. First, maneuverability –
a maneuverability that rivals the most remarkable flying critters -- birds and
insects. The ideal surveillance and reconnaissance mini-drone would be
able to “investigate” its environment with the resourceful maneuverability of a
flying insect or bird searching for its next meal.
Second, DARPA wants camouflage. It’s one thing to develop a mini UAV able
to engage in surveillance and reconnaissance activities. But, the
information obtained through surveillance and reconnaissance is most valuable
when the "subjects of observation" do not know they have been observed.
To accomplish this, a mini UAV must not be easy to see or, at least, easy to
identify.
So DARPA wants a UAV that (1) maneuvers like the most acrobatic of birds or
flying insects and (2) will not be identified as a UAV by an observer.
Amazingly, these are not conflicting objectives. In fact, animal-like
maneuverability and camouflage dovetail perfectly.
What DARPA calls
“biologically inspired” is, in robotic technology, referred to as “biomimetics”
or “biomimickry.” These terms describe a kind of revolution in the
concept and design of robots.
In the 1950’s, the sci-fi vision of robotic technology was both exotic and
strange. The technology of the future was envisioned and presented as
something completely different and contrary to our natural biological
surroundings. However, when technology confronted reality, we biological
organisms seem to have had the last laugh because we could (and still can) do a
whole lot of extremely useful things that our most sophisticated technology
cannot.
The jeep took a basic automobile and raised the center of gravity, increased
the size and scale of the automotive suspension system and produced spectacular
off-road performance for a machine with wheels. But the wheel, itself,
was limited. Every Rover we’ve sent to Mars ended its life when it got
stuck.
Human beings aren’t the strongest animal in the forest, but if just two of us
were with those Rovers on Mars, we’d have extended their useful lives by
getting them “un-stuck” in short order. Why? Because we have a
repertoire of movements and leverage that we can use to apply force in almost
any direction. The best of those early sci-fi ’bots looked high-tech but,
in fact, were functionally stunted.
When sci-fi was still dominated by those inhuman and unnatural versions of
mechanistic technology, a new technological methodology was, quietly,
born. “Biomimetics” was the first term used to describe the development
of technology designed to imitate and replicate the activities of biological
systems and organisms. Then, another term, “biomimicry,” was widely
adopted to describe any technology imitating (copied from) from nature.
But, in some contexts, biomimicry is more of a necessity than a choice.
If you want drones that work in a particular way, and the only known example of
such performance is a biological organism, you’ll either have to imitate the
organism or forget the project altogether. So, to get flying ‘bots that
maneuver the way flying insects and birds do, the ‘bots must be designed to
imitate the actual form and movement of these same creatures.
But, once you manage to replicate the form and functionality of a flying insect
or a bird in a UAV, how do you camouflage it? That's easy. Disguise
it as . . . a flying insect or a bird! In modern robotics, biomimckry and
camouflage often go hand in hand.
A little more detail. This form of camouflage is of a type called
“mimesis” or “masquerade.” The masqueraded object is quite visible to the
observer, but is designed to be mistaken for something else. That
“something else” would be of no particular interest to the observer. So,
when the mini UAV flies overhead, those who are being reconnoitered will think,
“It’s just a bird.”
An early attempt at this combination was the “Insectothopter.” Everyone
“knew” that the techno-savvy needed to build a robotic insect that actually
flew like the real thing was not available in the 1970’s. But the CIA did
it anyway. The laser guided Insectothopter was powered by a small
gasoline engine. Loud? Maybe, but if you’ve ever heard a real
dragon fly as it flies past your ear . . . . Well, you’d hardly notice
the difference.
But the Insectothopter never was “deployed” because of a stability issue.
For a mini UAV to be effective in surveillance and reconnaissance, the operator
had to be able to direct it to a defined target. With the Insectothopter,
a five mph cross-wind made such navigation impossible.
Only in 2011 was the next bio-inspired mini UAV unveiled. The “Nano
Hummingbird” or, technically, the “Nano Air Vehicle” (“NAV”) was the first
fully functional bird-drone designed and able to perform surveillance and
reconnaissance missions.
The “Nano Air Vehicle” project began in 2006 with Aero Vironment, Inc. working
under the direction of DARPA. If you know the history of the
Insectothopter, you can appreciate the significance of one of the DARPA project
requirements: the Nano Hummer must demonstrate the ability to hover in a
5 mph side-wind without drift of more than one meter.
Robo-Hummer was the first successful surveillance and reconnaissance bird
‘bot. But Robo-Hummer was designed in imitation of the hummingbird -- nature’s
version of helicopter. DARPA also wanted a bird ‘bot that would perform
more like a winged aircraft. The next bird ‘bot would be capable of
high-speed flight. But it would, also, have an appearance close enough to
that of a bird to achieve masquerade. In other words, the viewer would
mistake the mini UAV for a bird.
Florida’s Prioria Robotic responded with a mini UAV -- Maveric. In
terms of its field capabilities, Maveric is an amazing technical
achievement. This UAV can be carried, launched and operated by a by
a single person. The entire bird ‘bot is “bendable” and when not in
operation, is carried fully assembled, but folded up, in a 6 inch tube.
The operator simply (1) pulls it out of the tube, (2) powers it up, (3)
checks its sensors, and (4) launches it by throwing it into the air.
In flight, Maveric can reach a speed of 55 knots (63 mph). More amazing is
Maveric’s fully autonomous operation from launch to landing. How does it
know its target? It has a point-and-click feature that allows the
operator to find the target on a screen. Then, point and click on the
target. After launch, what does the operator do next? Often,
nothing.
Maveric is able to operate with full autonomy from launch to landing.
This ‘bot doesn’t require a human pilot to operate it in flight. The
operator need only define the mission – that is -- where to fly and where to
land. How does Maveric do it? Through the use of a system called
Merlin.
Merlin is Prioria’s proprietary processing platform. The Merlin “brain”
is carried and contained entirely on board Maveric during flight. The
“brain” processes images and operates vision-based controls – without the
assistance or necessity of a human operator. This makes Maveric the first
smart SUAS, “Small Unmanned Aerial System.” Of course, Maveric can
also be flown manually with a joystick while still providing the “pilot” with
autopilot assistance.
Of course, from a distance, Maveric looks like a bird. And, in many
environments, can be expected to be mistaken for bird. Its smart system
also provides a measure of stealth in terms of electronic transmissions.
Radio signals providing a live audio or video feed to the remote operator
are as common in surveillance drones as they are problematic. The
“problem” is radio waves. These radio transmissions create a detectable
signature that can disclose the presence and location of a surveillance
drone.
With Maveric, the operator can control the timing of transmissions to delay
broadcast of the live feed in sensitive locations – locations in which a radio
transmission might be detected and recognized. These transmissions aren’t
lost -- only delayed. Transmission will resume based on the
operator’s instructions.
Of course, some might question the usefulness of this radio stealth.
After all, Maveric’s use of radar and other navigational signals could just as
easily disclose its presence. But Maveric makes only very limited use of
any “other signals.”
How does it manage
that?
Merlin has some more magic. The Merlin “Brain” also engages in on-board
image processing with on-board vision based controls. In other words,
Maveric avoids obstacles and finds its target by sight. This 'bot
carries a self-contained vision-based navigation system, which allows it to
perform it mission without real time human assistance, while minimizing the use
of any electronic navigation systems that produce detectable radio signatures.
Of course, Maveric doesn’t look exactly like a bird. It has the
silhouette (shape and color) of a bird, but, unlike Robo-Hummer, Maveric
doesn’t have the bodily movements of a bird in flight.
Is
that an issue? Well, DARPA is working with the Army Research Laboratory
and the Maryland Robotics Center to produce yet another bird 'bot with the
flapping wings and the bodily movements of a real bird. It’s still in
development, but its name is Robo-Raven.
Mark Grossmann of Hazelwood, Missouri & Belleville,
Illinois
No comments:
Post a Comment