Anders Nelson (EE) [email protected] Mathew Wymore (CprE) [email protected] Kshira Nadarajan (CprE) [email protected] Mazdee Masud (EE) [email protected].

1

IRP PresentationMay 11-10: Autonomous UAV Competition

Anders Nelson (EE) [email protected] Mathew Wymore (CprE)[email protected]

Kshira Nadarajan (CprE)[email protected]

Mazdee Masud (EE)[email protected]

Client: Space Systems & Controls Laboratory (SSCL)Advisor : Matthew Nelson

2

Introduction

SSCL(Space Systems & Controls Lab)

Unmanned Aerial Vehicle

AUVSI Aerial Robotics Competition

3

Problem Statement

Aim: To participate in the International Aerial Robotics Competition(IARC) http://iarc.angel-strike.com/ Overall Challenge: To penetrate a building,

navigate through the corridors and completing tasks such as replacing a USB stick▪ Our specific challenge: To build a platform

capable of flying autonomously, stabilizing and avoiding obstacles

4

Functional Requirements

1.5kg Maximum Total Platform Weight Low Power

Battery Powered Capable of >10 minutes of flight time (20+

minutes reached with 1278g weight) Operational

Capable of onboard stability control (Generation of PWM signals from Pic )

Wireless base station communication▪ Wireless link capable of at least 42 meters

5

Functional Requirements (continued)

Expandable Potential for navigation in a GPS-denied

environment▪ Obstacle Detection Module using Laser Range

Finder▪ Support for computer vision system

Connectivity for manual remote kill switch Potential for remote autonomous commands Connectivity for wire-burn USB stick drop-off

6

Project Plan

Working with 2 other teams this semester Engr 467 – Platform Team worked with last

semester Engr 466 – Second Semester, Controls Team

for Project Weekly team and large group meetings

Team lead from Engr 467 leads group meeting and combines agendas from each team

Working closely with Controls Team to set API and sensors to their requirements

7

Full Design Approach

Quad-Copter Sensors

9 DoF IMU Laser Range Finder Camera Vision System

Control System Dedicated stability

controller Remote Kill Switch

and Manual Control Ability

8

Electronic Component

WiFi to base station for vision processing Two separate control modules

PIC dedicated to stability control Main controller for system control and

communication Internal and External Sensors

IMU - stability algorithms in PIC Altitude - detection through Sonar Laser Range Finder - collision avoidance and

navigation Cameras – finding room and USB as part of

challenge

9

Functional Decomposition

Hardware System Processing

Software System Software

Architecture Sensors System

Internal and External

Power System

10

Parts Selection

Laser Range Finder - >$2000 Hokuyo URG-4LX - 240º range : 2mm – 5m Obstacle Detection

Sonar – LV-Max Sonar-EZ 4 Depth Sensing Obstacle Detection, Control System

IMU – Analog ADIS 16400 9 Degrees of Freedom 12bits Resolution: +/- 300˚/s, +/-18g, +/-2.5 gauss

Gumstix Overo Fire COM TI OMAP3 720 MHz w/ 256 MB RAM Angstrom embedded Linux kernel Integrated USB, WiFi

PIC32MX795F512L 80MHz, 32-bit MCU

11

Midterm Challenges

Larger number of teams Controls Team: Vision and Navigation

Improved budget Re-evaluation of system design Arrival of some parts Communication Breakdown

12

Implementation

Manually flyable platform Table-top test system

Hardware Software Sensors Power

13

Hardware System

Gumstix SPI Ad-hoc WiFi w/ DHCP Gcc

PIC Sonar PWM SPI

PICstix

14

Software System

Overall system level software

Obstacle Detection Module

15

16

Obstacle Detection Module

int isLeftBlocked(); int isRightBlocked(); int isFrontBlocked(); int isBlocked(int index1, int index2);

17

Sensors System

Hokuyo URG-04LX Laser Range Finder Connection through software established Program developed for initializing and reading in data▪ Sub function of forward and side obstacle detection

Sonar Connects via USART serial in Consistent altitude measurements into PIC32

IMU Was not acquired, deemed a low priority▪ Gaui 330 by Controls team used to simplify controls▪ IMU still needed, but flight capable through stability system on

Gaui Were advised to concentrate on integration of other

parts

18

Power System

Lipo Battery 2 – 3cell pack combination will be used 11.1V, 6400mah, 20 C 320g

BEC voltage regulator 11.1V (ESC’s, 4 motors) 3.3V (PIC) 5V (Laser, Gumstix, Camera, Sonar,

IMU)

19

Power System (Continued)

11.1 V Lipo Battery3-cell

ESC’s

Motor Motor Motor Motor

Laser

Gumstix

Camera

Sonar

IMU

PIC

11.1 V

3.3 V

Voltage Regulator 5 V

Current Measure

20

Test Plan

SPI Oscilloscope, register watch, Gumstix

terminal Sonar

Oscilloscope, register watch Fixed position, yard stick, LEDs

Power Manual flight endurance test

21

Testing Results

SPI PIC and Gumstix run at different voltage

Sonar Sonar is accurate to 1 inch Ranges up to 44 inches

Power Flight time of almost 21 minutes With 1278g weight

22

Obstacle Detection - Testing

15 trials in each test caseSide True

Positives (%)

True Negatives (%)

False Positives(%)

False Negatives(%)

Left 100 0 14.28 85.71

Right 100 0 35.71 64.28

Front 98.4 1.6 28.571 71.42

Average 99.46 0.5 26.19 73.81

23

Future Work

Implement IMU with format for use by Controls team

Set up cameras with frame grabber Order designed PCB

Test all connections and functions Conduct stability tests with all items on

platform Implement navigation algorithm

Basic algorithm designed by Controls team Test runs on complete platform

Debug as needed

24

Lessons Learned

Importance of Communication Too many visions, too little time But I thought….

Full team from start Playing catch up Defining roles

Attention to detail Biting off more than you can chew

25

Questions?

26

Backup Slides

27

Hardware Options Scoring

28

Sonar Testing

29

PCB Schematic

30

Initial Schedule

31

Mid-Semester Schedule

32

Flight Tests

First successful flight http://

www.youtube.com/watch?v=SKbRr4Nj6V4

Demo of ESC shutoff at low voltage http://

www.youtube.com/watch?v=Hx0s8WKSEH8

Endurance Flight http://www.youtube.com/watch?v=jSsSb

n8tQgk