Colony

Investigating Scalability in a Robot Colony¶

Accepted by URO.

Development Notes¶

• Layout
  • Abstract
  • Research Question and Significance - Suresh
  • Project Design and Feasibility - Felix
  • Background - Eugene
  • Feedback and Evaluation - Greg
  • Dissemination of Knowledge - Kevin
  • Budget Research - Greg & Kevin
• Topic Ideas
  • Behaviors
    • Herding
    • Worker ants - carry things from one side to another
    • Chemical detection - detect chemicals
    • SLAM - making a map using wireless
    • Cooperative behaviors (pushing things)
  • Parts
    • Magnetic Encoders
    • Sharp IR sensor
    • Wireless
    • LCD (+ daughterboard)
  • Base redesign
    • Bump sensors - bump sensor ring
    • Buy several new bases
    • LCD mount + daughterboard
    • Battery placement
    • Wire management
• Research Question and Significance
  • The colony project seeks to develop behaviors that control a large number of robots working together in a team. Where one robot alone can malfunction and fail to accomplish a task a group of robots can complete goals even when individual robots break down. Past development has focused on improving the individual robot design. Now, with a relatively inexpensive platform, the goal is to create several robots. The project aims to create advanced behaviors for this large group, which the robots would not be able to perform individually.
  • Behaviors that have been developed already will be extended. What was once performed by a few robots will be performed by several. This includes robots moving together in geometric formations and also robots working together to surround a moving target. Previous formations were very simple, but with many robots more sophisticated maneuvers can take place. With more robots a colony can detect and surround its target in less time.
  • There are many real-world applications for colony based robotics. Landmines are a very serious problem in many parts of the world. A colony of robots searching for hidden mines can have greater coverage with a larger number of units. It also decreases the chances of missing mines.
    Robots can also be used to contain chemical spills. When it is too dangerous for humans to get involved or if the spill is in an unknown place a colony can search and surround the spill, preventing it from advancing and also characterizing the shape of the spill. The more units the colony would be able to locate a spill faster and also have better resolution when determining the shape of the spill.
    It is the goal of the project to facilitate research in multi-robot systems.
• Background Section
  • The Colony team is experienced, capable, and eager to continue working on the project. Felix Duvallet is (currently?) a junior Electrical and Computer Engineering major with a minor in Robotics. He has participated in the Colony project since its inception, and he has worked on numerous other robotics club projects. His work has centered around robot communication and localization. ... Aaron Johnson is a sophomore in Electrical and Computer Engineering. He has worked on several robotics projects through the robotics club, and he is researching snake robots in the Sensor Based Planning Lab in the Robotics Institute. ... Ryan Kellogg is a junior majoring in Electrical and Computer Engineering and Biomedical Engineering. He has worked on the Colony project for three years, and he has done research in the NanoRobotics? Lab in the Robotics Institute. ... Iain Proctor is a sophomore in Cognitive Science and Computer Science. ... Suresh Nidhiry is a sophomore Computer Science major... Eugene Marinelli is a freshman majoring in Computer Science and Electrical and Computer Engineering. This year he has worked on behaviors for the Colony. He also does robotics research for the Claytronics group in the Computer Science Department and programs for the Robocup small-size team. Kevin Woo is a freshman Electrical and Computer Engineering major. ... James Kong is also a freshman in Electrical and Computer Engineering. ... Gregory Tress is a freshman Electrical and Computer Engineering major ... Abhishek Krishnan is a freshman in Computer Science. ...
• Dissemination of Knowledge
  • We plan to pass on the knowledge and technology created by the Colony Project in a myriad of ways. We will publish the code and the workings of the colony on the project’s website. By allowing others to see and manipulate the code, they will be able to create their own low cost colony without the hassle of creating the base technology from scratch. This will in turn allow them to focus on researching emergent behaviors.
    The microcontroller and the library for the board will be sold on Botric’s store so that other robotics projects have a solid technological base to work from. Furthermore, the board will be used in the Fun With Robots StuCo?. Both of these will allow those working on robotics projects to begin creating, learning, and researching with robots without the need to build a technological base. This will allow everyone to progress toward their goals much more quickly.
    The team will also publish the results of the emergent behaviors discovered with a larger colony on the project website. All findings will also be presented at the Meeting of The Minds.
• Feedback and Evaluation
  • The status of the project is evaluated regularly at our weekly project meetings. We observe which objectives we have accomplished, which ones are still in progress, and how these objectives have contributed to the overall goals of the Colony project. We also use this information to decide what new goals should be added to the project in terms of hardware, software, and design features and what new equipment and/or funding will be necessary to reach these goals. Professor Howie Choset is our advisor and attends various project meetings. Professor Choset is able to offer feedback from a professional perspective, which helps to guide the development of the Colony, and which helps us to solve some our technical issues.
• Budget stuff
  • Possible items:
    • LCD modules 19.95 ea (previously asked for 0)
    • Wireless modules, 15 ea (previously asked for 0)
    • Rest of microcontroller boards, 100 ea (previously asked for 8)
    • IR ring, 19 ea from acroname, 5ea from club, (previously asked for 5 sonars for testing)
  • Price Estimates from Colony II
    • Batt $10
    • Base $35
    • Cereb $95
    • IR $12
    • Wireless $40
    • Pyro $60
    • "Beacon"/BOM $16.50
    • Bump $5.92/robot
    • Total: $274
  • Complete old budgets:
    • Spring 06: Total: $1160
      • New microcontroller boards, 8x$100
      • Sonar sensors, 5x$20
      • Encoders, 16x$10
      • Mech. Parts $100
    • RBOM spring 06: Total: $1020
      • 12 boards, each $85, as follows;
      • Atmels 12x$8
      • FPGA 12x$20
      • IR emitters/detectors 12x$15
      • Board fab 12x$6
      • Other elec 12x$5
      • RF tx 12x$13
      • RF rx 12x$18
    • Spring 05: Total: $1010
      • Math coprocessor boards 12x$60
      • Bump sensors 4x12x$1.66
      • Pyro 4x$52.50
    • Emergence Spring 05: Total: $830
      • Wireless Boards $50 (for 12 robots worth)
      • Tx/Rx Pairs 12x$60
      • Other 12x$5
    • Spring 07 Estimates (temporary)
      • 20 robots ~= $3220 total
      • 16 robots ~= $2180 total
      • 12 robots ~= $1140 total
      • estimates take into consideration prior budgets and parts
        • estimates include LCD, Zigbee, Firefly, IR Rangefinder, RBOM, Encoder
        • estimates do not include Mechanical Parts
        • see attached spreadsheet

Files¶

• Final Version: Colony_F06_SURG.pdf
• Budget Calculator: Colony_-_F06_Budget_Calculator.xls