Tech Briefs Magazine - May 2021 - 34

Aeronautics

Navigation Algorithm Enables Drones to Learn Acrobatic
Maneuvers
Autonomous quadcopters can be trained using simulations to increase their speed, agility,
and efficiency.
University of Zurich, Switzerland

ilots have long used acrobatic maneuvers to test the limits of their airplanes.
The same goes for flying drones. Professional pilots often gauge the limits of
their drones and measure their level of
mastery by flying such maneuvers.
Working with Intel, researchers have
developed a quadrotor helicopter, or
quadcopter, that can learn to fly acrobatic maneuvers. While a power loop or a
barrel role might not be needed in conventional drone operations, a drone ca-

pable of performing such maneuvers is
likely to be much more efficient. It can
be pushed to its physical limits, make full
use of its agility and speed, and cover
more distance within its battery life.
The researchers developed a navigation
algorithm that enables drones to autonomously perform various maneuvers using
nothing more than onboard sensor measurements. To demonstrate the efficiency
of their algorithm, the researchers flew
maneuvers such as a power loop, a barrel

roll, or a Matty Flip, during which the
drone is subject to very high thrust and
extreme angular acceleration.
At the core of the algorithm lies an artificial neural network that combines input
from the onboard camera and sensors
and translates this information directly
into control commands. The neural network is trained exclusively through simulated acrobatic maneuvers. This has several advantages: Maneuvers can easily be
simulated through reference trajectories
and do not require expensive demonstrations by a human pilot. Training can scale
to a large number of diverse maneuvers
and does not pose any physical risk to the
quadcopter.
Only a few hours of simulation training
are needed for the quadcopter to be ready
for use, without requiring additional finetuning using real data. The algorithm uses
abstraction of the sensory input from the
simulations and transfers it to the physical
world. The algorithm learns how to perform acrobatic maneuvers that are challenging even for the best human pilots.
Drones used for search and rescue missions or for delivery services will benefit
from being able to cover long distances
quickly and efficiently.
For more information, contact Prof. Dr.
Davide Scaramuzza at sdavide@ifi.uzh.ch;
+41 44 635 24 09.

A quadrotor performs a Matty Flip. (Image: Elia Kaufmann)

Predicting In-Flight Air Density for More Accurate Landing
An algorithm runs onboard a vehicle, providing important real-time data to aid in steering
the craft.
University of Illinois at Urbana-Champaign, Urbana

can run onboard a vehicle, providing important real-time data to aid in steering
the craft, particularly during the crucial
entry, descent, and landing stage.
The algorithm starts without knowing
anything about the air density. It gathers
data from accelerometers and gyroscopes available on any vehicle to gather
data and combines it with prior knowledge about maximal rate of accelera-

www.techbriefs.com

n the final few minutes of a spacecraft
landing, it is moving at hypersonic
speed through many layers of atmosphere. Knowing the air density outside
the vehicle can have a substantial effect
on its angle of descent and ability to hit a
specific landing spot. But air density sensors that can withstand the harsh hypersonic conditions are uncommon. Researchers developed an algorithm that

Cov

ToC

tion to obtain a time-varying estimate of
air density. It learns and changes its estimations onboard, based on the input
data it receives.
The researchers used data acquired
from the entry, descent, and landing of
the Phoenix lander - a Mars science
probe - representing the last 220 seconds (the ballistic phase) until parachute deployment. The craft cannot be
Tech Briefs, May 2021

http://www.techbriefs.com http://www.abpi.net/ntbpdfclicks/l.php?202105TBNAV

Tech Briefs Magazine - May 2021

Table of Contents for the Digital Edition of Tech Briefs Magazine - May 2021