Robotic Space Colonization Projects
SEI's Robotic Space Colonization team is working to develop autonomous robotic systems for the construction of a lunar or Martian habitat. This team objective is aligned with NASA's Vision for Future Space Exploration.
Spring 2008: TSGC Design Challenge - Semester II
The team is continuing with the Fall-2007 objective, which is the development of a team of robots that will autonomously lift and transport a long construction object. This semester the team has focused on the design of cooperative-control laws for a two robot team, the design of the central-PC software architecture, and systems integration. The semester goal is a hardware demonstration of the autonomous system.
Semester reports can be viewed here:
- Report 1
- Report 2
- Report 3
Fall 2007: TSGC Design Challenge - Semester I
The Robotic Space Colonization Team is researching and implementing in hardware a system of autonomous robots that can perform a cooperative-lifting task with two robots. A cooperative-lifting or construction task is a step towards the development of autonomous robots that can assemble a Lunar or Martian space habitat for human occupation. The project will be a technology demonstration that will focus on the development of cooperative control laws for two robots. A major step towards control law development is the ability to estimate inertial position and velocity of the robots. A Kalman-Filter will be implemented to estimate robot positions using inertial measurements of the robots from an overhead camera, local position measurements from onboard wheel-velocity encoders, and a representative model of the robot equations of motion. Path-planning and navigation algorithms will drive the robots to a desired location and orientation. Formation-control or relative-control laws will direct the robots to achieve and maintain desired positions in order to perform the cooperative-lifting task.
The current project is related to the Advanced Robot Technology Topic through the continued investigation of robot autonomy to minimize human interaction and intervention throughout habitat construction tasks. Our main goal is to develop the ability to cooperatively lift and transport a large construction object using two robots. We have assumed that the construction object is a long object (with a length much greater than the longitudinal cross-sectional area) that can be grasped on each end by a robot. We believe that meeting the stated objective is the next step in developing robot autonomy for habitat construction. Initially, we will design the necessary cooperative control laws to lift and transport a flexible object. As we improve our control-law development, we plan to demonstrate the project objectives for a rigid construction object. More information can be found on the main page.
Spring 2007
The initial phase of the project, Phase I, focused on the development of a system of autonomous robots that could transport small construction objects from lunar or Martian landing sites to a habitat construction site. Fixed or known robot highways were a key simplification in this first phase of the project. In order to accommodate rapid development, the team implemented a centralized-control system to track robot locations and allocate tasks to the individual robots. During this phase of the project, several critical subsystems were developed including wireless communication between the robots and the central PC, motion-control commands using infrared line-following techniques, and navigation algorithms using a weighted-graph representation of the design space.
Fall 2006: Lego Mindstorm
This project was an inter-team competition of three teams in order to give us the opportunity to work with a simple robotic system. Each team was given a Lego Mindstorms NXT kit and either used the Lego Mindstorms software or RobotC for programming of a fully autonomous robot. The requirements of the competition included navigating autonomously through a maze, locating a burning candle, and extinguishing the candle. The teams used touch, light, and ultrasonic sensors for navigation and location of the candle, and a catapult contraption for extinguishing. The main challenge of the project was programming the software so that the robot was able to overcome any potential problem in order to have a fully autonomous system with no human interaction.
Fall 2006: Digital Signal Processing
Digital Signal Processing This project was another inter-team competition with two teams that focused more on the hardware aspect of robotics. The objective was to design and build a remote controlled system to collect 'dry' laundry and play 'wet' laundry on clothesline within a three minute time limit. Each team took a different approach to this challenge. The first designed a telescoping arm using a car antenna with a claw like end to reach the bandanna. The second team used a two link mechanism controlled with a motor as an arm and wheels capable of going over a 2x4. Challenges included attaching the motor to the wooden arm and wheels and designing a effective claw mechanism.