Spring 2011 - Winch Controlled Balloon Lifting
with Movable Robotic Feet

This research was conducted with Daniel Nevius and Jenny Zhang under the guidance of Professor Robert Howe and Barry Griffin as a 91r independent research class at Harvard University.

The purpose of this project was to develop a winch design and working prototype for a balloon powered technology that harnesses the lift and buoyancy provided by inflated helium balloons to lift and move heavy payloads.  The system consists of a suspended helium balloon connected to winching systems on top of movable robotic platforms.  It could be implemented on land with helium balloons or under water with buoys.  It has real world applications for members of the marine industry, governmental organizations, and the United States Navy.  For example, the system could be used for deep-sea mining, oceanic clean up, or maintenance of fish farms.

The primary task of the system is to move payloads to a desired location by winching the balloon and moving the robotic feet.  A remote computer takes a payload position as user input and sends coordinates to the robots over over an 802.11 wireless connection.  An Arduino Uno on each robot receives the coordinates and commands the robot to move to the desired location with the desired winch position.
Picture

Specifications

Processor- Arduino Uno with SparkFun WiFly Shield
Sensors - Quadrature encoders for drive motor and winch motor
Software - Arduino programming language (basically C)
Power - Two AAA and four AA rechargeable batteries per robot
Drive Type - Omni wheels for two axis motion only

Documentation

INTRODUCTION
The main design goal for this project was to do concept development work and winch design for a  working  prototype  of  a  Winch  Controlled  Lifting  System  with  Moveable  Robotic  Platforms.  Simplicity  was  a  primary  goal  because  the  system  is  only  a  proof  of  concept  and  time  and resources were limited.  The specifications were as follows:
  • There will be a maximum of four meters between robots and the top of the balloon will be no higher than three meters.
  • A 3-foot diameter balloon will be used.
  • The payload will be less than 1 ounce.
  • 1 cm accuracy will be achieved for payload positioning.
  • There will be no sensor feedback.  Only encoders will be used.
  • The system will be controlled from a remote computer.

ROBOT DESIGN
The robotic platforms were designed and built with the basic principles of simplicity, functionality, and practicality in mind.  The robots were built using basic VEX Robotics Design System parts, hardware, and pieces of machined/cut metal.  One robot consists of a rectangular metal chassis, a basic  interconnected  four-wheel  drive system  regulated  by  two  optical  shaft  encoders,  and  a rectangular  overhead  plate  on  which  the  winching  system  is  mounted.    Once  input  is  received from a computer program, the robot moves by completing a series of strict forward and side-to-side  movements  (as  if  it  were  traversing  a  checkerboard)  in  order  to  reach  any  given  location. Optical shaft encoders, one for each axis of movement (x an y), measure the distance travelled by the robot.  
Picture
WINCH DESIGN
Each  of  the  robotic  feet  has  a  winch  attached  to  the  top  used  to  manipulate  the  lengths  of  the ropes  attached  to  the  balloon  in  order  to  position  it  between  the  feet.    The  winch  can  be subdivided into four components: the motor, drum, control, and feedback.
Picture
ONBOARD SOFTWARE
  • WiFly Shield receives wifi packets and send them to the Arduino Uno
  • Arduino program runs on board startup, parses the incoming packets, and sends motor commands
COMMUNICATION
  • Two way communication betweenremote computer and Arduino Uno to send commands and sensor data over 802;11 wireless network
  • Arduino on the Arduino Uno parses packets and sends drive commands to the motors through PWM signals
  • Extensible packet structure to easily incorporate more sensors and motors
Picture

Media

 
MORE INFORMATION

Paper

Final Presentation