Target Drone Platform
Hardware
Traxxas Slash VXL
Traxxas is known for fast radio controlled cars and
that is what system requirements demanded.
After tracking a projectile, the target drone has
only a small window of time before projectile
impact. Calculations revealed that the system
required a vehicle that would accelerate quickly to
a high speed as fast as possible, while being able
to carry a heavy load. The Traxxas Slash VXL is the
perfect vehicle for the system. It can exceed
50MPH and was big and powerful enough to carry
the load requirement.
FEATURES ON 1/16 SLASH 4X4 VXL:













Ready-To-Race® with included 6-cell NiMH Power Cell battery and charger
30+mph with included NiMH battery, 50+mph with additional NiMH battery
TQi 2.4GHz Radio System
Revo suspension technology for superior handling and control
Full-time 4WD
Velineon® 380 brushless motor—largest in class
Waterproof electronics
Torque-control™ slipper clutch
Sealed differentials
Realistic all-terrain tires with foam inserts
Traxxas High-Current Connector
Metric hex hardware
Rubber sealed ball bearings
ArduPilot Mega (APM)
This Arduino-compatible autopilot board by DIY
Drones was used to win the autonomous vehicle
competition organized by Sparkfun. Ardupilot is a
fully programmable autopilot that requires a GPA
module and IMU sensors to create a functioning
Unmanned Ground Vehicle. The hardware and
software and all open source and can be
programmed with the Arduino IDE. APM can be
used for airplanes, rotary wing aircraft, or ground
vehicles, depending on the firmware loaded.
Features on the ArduPilot Mega:














Designed to be used with autonomous aircraft, quadcopters and helis
Based on a 16MHz Atmega2560 processor.
Built-in hardware failsafe that uses a separate circuit (multiplexer chip and ATMega328
processor) to transfer control from the RC system to the autopilot and back again.
Includes ability to reboot the main processor in mid-flight
Dual-processor design with 32 MIPS of onboard power
Supports 3D waypoints and mission commands (limited only by memory--approximately 600700 waypoints with current code, but can be adapted for more)
256k Flash Program Memory, 8K SRAM, 4K EEPROM
Comes with a 6-pin GPS connector (EM406 style).
Has 16 spare analog inputs (with ADC on each) and 40 digital input/outputs to add additional
sensors
Four dedicated serial ports for two-way telemetry (using optional XBee modules) and
expansion
Can be powered by either the RC receiver or a separate battery
Hardware-driven servo control, which means less processor overhead, tighter response and no
jitters
Eight RC channels (including the autopilot on/off channel) can be processed by the autopilot.
LEDs for power, failsafe status, autopilot status
ArduPilot Mega IMU Shield
The APM IMU Shield is designed to fit on top or
bottom of the ArduPilot Mega board, creating a
total autopilot sulotion when a GPS module is
attached. This board features a large array of
sensors needed for Unmanned Autonomous
Vehicles (UAV) and Robotics applications,
including three axis angular rotation and
acceleration sensors, absolute pressure and
temperature sensors, and 16MBits data logger
chip.
Features on IMU Shield:













Dual 3.3V Regulator!!! (One dedicated for analog sensors)
Relay switch for cameras, lights or payloads.
12-bit ADC for better Gyro/Accel/AirSpeed resolution.
Built-in 16MB Data Logger (The Black Box).
Piano DIP switch for servo reverse or user customizable.
Built-in FTDI, making the board native USB.
Dedicated Modem/OSD port.
I2C Port with incoming "Daisy Chain board" allowing you to build sensor arrays.
Two user-programmable buttons.
10-Bit analog expansion ports.
Reset button.
Optional "Through Hole" voltage dividers (Easy to solder).
Tons of Status LEDs.





New vibration resistance Invensense Gyros (Triple Axis).
Analog Devices ADX330 Accelerometer.
AirSpeed sensor port (optional, sold separately).
Absolute Bosch pressure sensor and temp for accurate altitude
Weight: 0.5 oz; 13 g
MediaTek MT3329 Global Positioning System (GPS)
A state-of-the-art 66 channel MediaTek MT3329
GPS Engine, perfect to aid in positioning of the
target drone. Its high sensitivity is perfect for urban
performance. It includes an adapter to easily be
added to ArduPilot Mega’s IMU Shield.
Features on MediaTek MT3329 GPS:
 Based on MediaTek Single Chip Architecture.
 Dimension：16mm x 16mm x 6mm
 L1 Frequency, C/A code, 66 channels
 High Sensitivity：Up to -165dBm tracking, superior
 urban performancesPosition Accuracy：< 3m CEP











(50%) without SA (horizontal)
Cold Start is under 35 seconds (Typical)
Warm Start is under 34 seconds (Typical)
Hot Start is under 1 second (Typical)
Low Power Consumption：48mA @ acquisition, 37mA @ tracking
Low shut-down current consumption：15uA, typical
DGPS(WAAS, EGNOS, MSAS) support (optional by firmware)
USB/UART InterfaceSupport AGPS function ( Offline mode : valid up to 14 days )
Includes a molex cable adapter, 5 cm.
Includes the new basic adapter.
Weight: 0.3oz; 8 g
Target Drone Controller
Zotac ION Intel Atom D525 Motherboard
The Target Drone Controller will perform
all computations regarding projectile
detection, projectile tracking, and
projectile avoidance. The Zotac ION
includes an Intel Atom D525 1.8 GHz Dual
Core CPU with onboard NVIDIA ION
Graphics. This allows for high
performance for a single core and the
option to parallize, while staying light
weight.
Features on Zotac ION Intel Atom D525 Motherboard:

Microsoft Windows 7 Premium











CPU: Intel Atom D525 (1.8GHz, Dual-Core)
Hyper-Threading
Chipset: intel NM10
Memory: DDR3 800 (4GB Max)
1 PCI Express x4 expansion slot
1 x Mini PCI Express (support WiFi module)
2 x SATA 3.0 Gb/s
Next-Generation NVIDIA ION Graphics (CUDA Ready)
1 x HDMI
4 x USB 2.0, 2x USB 3.0
Onboard DC-DC power supply
 Dimensions: 6.7” x 6.7”
Crucial 4GB (2 x 2GB) DDR3 Memory
The Zotac ION’s maximum memory capacity is
4GB of memory. The Target Drone is designed
to work in real time, so it is crucial to optimize
the Target Drone Controller for maximum
performance. Maximizing the memory
capacity does just that.
Features of Crucial DDR3 Memory:






240-Pin DDR3 SDRAM
Model: CT2KIT25664BA1339
Voltage: 1.5V
ECC: No
Unbuffered
Dual Channel Kit
Toshiba 120GB 5400 RPM 2.5” HDD
A solid state drive was preferable with its light
weight and increased performance, but due to
mini PCI slot not supporting SATA, it could not be
used. This Toshiba HDD fulfilled our
requirements with a 5400 RPM motor that gives
it up to a 12% performance boost over 4200RPM
drives. The SATA150 interface and 8MB cache
also provides a high speed pathway for data.
Features of Toshiba HDD:




Model: MK1234GSX
SATA 1.5 Gb/s
8MB Cache
Average Seek Time of 12ms



Average Latency of 5.55ms
5400RPM
2.5” Form Factor
LIPO (Lithium Polymer) Batteries
Lipo batteries are a type of rechargeable
battery that has taken the electric RC world
by storm. They are light weight, which is a
major concern for the Target Drone. They
also have large capacities, meaning they hold
lots of power in a small package. They also
have high discharge rates to power the most
demanding electric motors, like the Target
Drones motor.
Software:
EMGU CV 2.3.0
Emgu CV is a cross platform .Net wrapper to the Intel
OpenCV image processing library. Allowing OpenCV
functions to be called from .NET compatible languages
such as C#, VB,VC++, IronPython etc. The wrapper can be
compiled in Mono and run on Linux / Mac OS X. C#’s
modern object oriented language and rich class library
allows rapid development, while OpenCV is a great
opensource library of programming functions for real
time computer vision.
Microsoft Windows 7 Premium
Windows 7 is running of the Target Drone Controller.
Windows is a reliable operating system that offers all the
features necessary to Control the Target Drone. The
Windows 7 Operating System is also necessary to run the
Microsoft Kinect SDK 1.0 & Microsoft Robotics Developer
Studio 4. Implementing these software products will
dramatically improve development time and Target Drone
performance.
ArduRover
Code for ArduPilot
Microsoft Robotics Developer Studio (RDS) 4
RDS 4 provides a wide range of support to
help make it easy to develop robot
applications. RDS 4 includes a programming
model that helps make it easy to develop
asynchronous, state-driven application. RDS 4
provides a common programming framework
that can be applied to support a wide variety
of robots enabling code and skill transfer.
RDS 4 includes a lightweight asynchronous
services-oriented runtime, a Concurrency and
Coordination Runtime (CCR), Decentralized Software Service (DSS), and a
simulation engine and tools.
Kinect for Windows SDK v1.0
The Kinect for Windows SDK enables
developers to create applications that support
gesture and voice recognition, using Kinect
sensor technology on computers running
Windows 7. This software gives enables the
use of the Kinect RGB and Depth cameras to
detect, track, and make the path of projectiles
fired at the Target Drone.
System requirements:
Supported operating systems: Windows 7
Windows Embedded Standard 7
Hardware Requirements
32-bit (x86) or 64-bit (x64) processor
Dual-core 2.66-GHz or faster processor
Dedicated USB 2.0 bus
2 GB RAM
A Microsoft Kinect for Windows sensor
Software Requirements
Microsoft® Visual Studio® 2010 Express or other Visual Studio 2010 edition
NET Framework 4.0
Projectile Detection / Tracking Algorithms
The projectile detection and tracking is
implemented for the Target Drone Platform for
each frame, as it is processed. To detect the
projectile, color threshold is used with the
current threshold values set for detection of the
color orange. To track the projectile, image
moments are used. From the moment, the X
and Y coordinate of the center point of the
tracked object is found and then used to find
the depth coordinate (Z) at the position of the
object. The coordinates are then sent to the
appropriate service and used for the projectile
motion calculations.
Projectile Motion / Collision Algorithms
It is necessary to find the vector of the projectile
in order to determine if the projectile will collide
with the Target Drone. The computer vision
algorithms determine points in space where the
projectile is located with timestamps. Using these
points and times the projectile motion equation
determines a parametric equation for the
projectile. This equation is used to predict the
projectiles path. A Sphere is used to represent
the Target Drone. If there is an intersection with
the sphere and the parametric equation
representing the projectile a collision is eminent.