COLONY Introduction September 11, 2009

COLONYCOLONYIntroductionIntroduction

September 11, 2009September 11, 2009

2

Outline

1. Motivation

2. Brief Introduction to the Robots

3. Recent Colony Work

4. Current Research

5. Administrative Things

3

Motivation

Create a colony of robots that does cool stuff

Interesting research platform– Emergent behaviors– Robotic cooperation– Multi-agent interaction– Distributed algorithms– Simultaneous Localization

and Mapping (SLAM)– …

4

Why Colonies?

Many successful organisms organize themselves into groups

5

Why a Colony of Robots?

Robustness– Single robot + single failure = game over– Colony of robots + single failure = one robot down

Distributed intelligence and sensing– One robot cannot be everywhere at once– Colonies can collect and communicate data across distant

points within an environment

Collective behavior– Cooperation between robots to accomplish complex tasks

Robots are awesome– More robots are more awesome

66

Goals

Many low-cost robots

Open, usable platform– Capable hardware– Robust open-source code base

Research– Multi-robot applications– Distributed algorithms– Emergent behaviors

Fun Stuff

7

Brief Colony History

Project started in 2003 by Steve Shamlian Robots

– Firefly I, II– Dragonfly

BOM, BOM 1.5 Many research grants

– URO - Over 10 Small Undergraduate Research Grants– Ford Motor Company– Robotics Institute

2 Papers / Conferences– NCUR – National Conferences on Undergraduate Research– AAAI – Association for the Advancement of Artificial Intelligence

8

Active Members

Project Leaders– John Sexton– Austin Buchan– Chris Mar

Current Members– Jimmy Bourne– Megan Dority– Emily Hart– Rich Hong– Abe Levkoy– Evan Mullinix

Distinguished Alumni– James Kong– Kevin Woo

– Bradford Neuman– Nico Paris– Ben Poole– Justin Scheiner– David Schultz

9

Robots

1010

Dragonfly

1111

Dragonfly – Front View

Front IR Rangefinders

1212

Dragonfly – Back View

Side IR Rangefinders

1313

Communication

Bearing and Orientation Module (BOM)

– Localization sensor– IR emitter/detector array– Relative angle

measurements to other robots

1414

Communication

XBee wireless module– 30m indoor range / 100m outdoor range– IEEE 802.15.4 (ZigBee)– 2.4 GHz– Low-cost, low-power– Open industry standard

15

16

Recent Colony Work

17

Behaviors

Individual and multiple robot behaviors Simple local interactions can yield complex

global actions– Emergent behavior

Use sensor data to control actions

18

Behaviors

Past Projects:– Obstacle Avoidance– Cooperative Maze Solving– “Follow the Leader”– “Marching Band” 1.0

19

Autonomous Recharging

Battery charge is limited

Recharging batteries is a pain

Let the robots charge themselves!

20

Mapping

Autonomous wall following

Obstacle detection using rangefinders

Position estimation using encoders

Wireless data collection

21

Simulator

Test robot behaviors quickly

Develop behaviors independently of hardware problems

Develop code even when robots are out of reach

Simulating 100 robots is a lot cheaper than having 100 robots

22

Diagnostic Station

Inconsistent sensors are consequence of inexpensive colony

Calibrate and characterize sensors for every robot

Automate colony maintenance

23

Current + Future Work

Formation Control (SURG Grant)– Description

Make robots form and maintain a spatial relationship Stay in formation while moving as a group

– Applications Cooperative Mapping Cooperative Object Manipulation “Marching Band” 2.0 Distributed Search Algorithms

<Your Idea Here>

24

FAQ

25

FAQ

Why should I join Colony?– Experience with robots– Learn about all phases of research

Proposals (i.e. SURG) Robotics (design, programming) Presentation

– Awesome long-term project Experience with motivated, talented team Exposure to embedded systems, sensors, wireless

communication, mobile robotics

26

FAQ

Do I need to know (skill) ?– No! – But you are expected to take an active role in

learning– Club members will provide assistance

How much time will this take?– Expect at least 3-10 solid hours per week– The more you put in, the more you get out of it

27

FAQ

How do I get started?– Make sure we can contact you

Get added to the email list:[email protected]

Email [email protected] to get added

– Attend work meetings First three weeks of meetings will be geared towards

getting you up to speed

28

AdministrativeThings

29

Meetings

Project Meetings– Fridays, 4:30pm. NSH 1109

Cookies!

– Everyone. Every time.– Status updates, administrative matters

Work Meetings– Monday and Wednesday, 6:30pm – 8:30pm in the

Club Room

30

Work Meetings

First few weeks are special!– Hands-on labs to bring you up to speed

Labs will be released online– Can work independently if unable to make work

meetings– Members will be in the Club during work meeting

hours

Bring your laptops!– Available computers are limited

31

Colony Introduction Schedule

Lab0 - Dance Competition– [Release date: 9/11/09] [Demo date: 9/18/09]

Lab1 - Dead Reckoning– [Release date: 9/18/09] [Demo date: 9/25/09]

Lab2 - Hunter-Prey– [Release date: 9/25/09] [Demo date: 10/2/09]

32

Things You Should Look At

C Programming Tutorial– http://www.eskimo.com/~scs/cclass/notes/top.html– Chapters 1 – 3– Don’t worry about compiling; we’ll show you that

WinAVR– Compiler, linker, loader package for programming our robots

Programmer’s Notepad– Code (text) editor bundled with WinAVR for convenient programming

SVN – Subversion– Source control

Redmine– Wiki– Task management system

33

Advisor

George Kantor– RI Project Scientist– Teaches several RI classes– Controls, sensor networks, …– Knows a thing or two about robots– Busy guy

34

Important Emails/Web

Colony list– [email protected] – new people

Project Leaders:– [email protected]– [email protected]– [email protected]

Web:– www.robotcolony.org