project manager: gabe louthan - george fox university · 2020-05-02 · background specifications...
TRANSCRIPT
Ad-hoc Wifi
LIDAR-Lite
v2 Sensors
USB Output BEC
Regulator
IRIS Drone
I2C
Raspberry Pi 3
IRIS Battery
USB
(MAVLink protocol)Radio Control
The increased usage of Unmanned Aerial
Vehicles (UAVs) has also brought an
increase in the number of environments that
drones are expected to be able to fly in.
These environments include areas low to
the ground that are likely to be crowded with
a variety of obstacles. Flying a drone in such
an environment is a difficult task, and
crashing a potentially very expensive UAV is
an all too common experience most would
probably like to avoid.
Proposal
System Housing:
The 3D printed sensor and Raspberry Pi housing fits snugly around
the body of the 3DR IRIS. The housing is designed to hold the Pi
under the lid and mount four sensors 90 degrees apart. It was
designed to minimize weight and allow the user easy access to the Pi.
Final Design
Drone Anti-Collision SystemProject Manager: Gabe Louthan
Technical Manager: Bryan Neufeld
Engineering Team: Keiko Fujii, Alex Spivey, Denis Yablonsky
Marketing Team: Keenan O’Hern, Thaddeus Hughson
Advisors: Dr. Robert Harder, Dr. Gary Spivey
Figure 2: SOLIDWORKS Model of the Housing Assembly
Background
Specifications
Figure 1: Garmin LIDAR-Lite V2 Laser
Distance Sensor
Senior Design
Keenan O’Hern, Alex Spivey, Gabe Louthan,
Bryan Neufeld, Keiko Fujii, Denis Yablonsky
Figure 5. Drone Collision Avoidance
PurposeIn order to make crashes less common, an
obstacle detection and avoidance system is
proposed. Using LIDAR-Lite sensors from
Garmin, the system will detect obstacles in
a remote controlled UAV’s path, and then
take control from the user in order to move
the UAV off a potentially perilous path, and
back to a point of known safety. The
purpose of the project is simultaneously to
fulfill a known need in the drone community
and to give Garmin a way to promote the
use of their LIDAR-Lite sensors in a new
market.
We propose a system of LIDAR-Lite sensors
connected to a Raspberry Pi, in a housing
system that could be mounted on a 3DR
IRIS drone. The Pi will be loaded with
software that will control the sensors and
take control of the drone’s flight if an
obstacle is detected, enabling it to avoid a
collision.
The system needs to support an array of
LIDAR-Lite v2 sensors, override the user if
an obstacle is sensed, and avoid the
obstacle. The design needs to weigh less
than 150 grams so that a 3DR IRIS can
support it while in flight. The system’s
software will process data from the sensors,
and make a decision about the current flight
trajectory. It needs to interface with the UAV
flight management system. The design’s
cost needs to be minimized, as hobbyists
are the target audience.
Figure 4: System Diagram
Electronics:
The Raspberry Pi runs software that interfaces with the Pixhawk, the
IRIS’s flight controller, via the MAVLink protocol. The software also
interfaces with the LIDAR-Lite sensors. When a reading is required of
a sensor, the sensor is powered on, and a reading is communicated
via I2C, and then the sensor is shut off. The Pi is powered by the
IRIS’s battery with the assistance of a USB output BEC regulator.
This eliminated the need for an external battery to power the Pi.
Figure 3: SOLIDWORKS Assembly of the System Housing
Collision Avoidance Algorithm:
The collision detection software records GPS
locations where the drone has been. When an
obstacle is detected by the sensors, the
algorithm determines a previous location to go
back to, and then communicates with the flight
controller, taking control of the drone and
sending it back towards a safe location. After
two seconds, control is returned to the user.
Team
Special thanks to Dennis Corey and Bob Lewis of Garmin and
Bruce Cleveland of Bend Poly
Acknowledgements