Project Tasks

Initial State (Tane)

• Goal: Build map at origin to account for LIDAR blindspot
• Dependencies: spin, pose_ctrl/A*
• Assumptions: obstacle free placement
• Drive 0.6 meters forward
• Spin 180 degrees CW
• Drive 1.2 meters forward
• Spin 180 degrees CW
• Drive 0.6 meters (back to origin)

Frontier Exploration (Andrew/Shawn)

• Goal: Explore environment collision-free until termination condition is met
• Dependencies: Occupancy grid, spin, astar, pose_ctrl
• Assumptions: known total map area
• Tradeoff between accuracy and speed
• When/how to spin to increase chance of object detection
• Idea 1: Choose furthest known point in straight line, go to point, repeat, if furthest point below threshold, turn 90 degrees and repeat
• Idea 2: Continually pick furthest known point
• WORKING NOW:
  • Pick the furthest 0 in the occupancy field and set that as cmd_nav goal
  • Need to publish this continuously as that is how the Navigator.py run_navigator() function works
  • Need to find out why some zeros are outside of current map? Wrong logic? Maybe pick a zero with other zeros bordering it...

Command Center Visualization (Allen)

• Have supervisor modes be visible on the robot
• Add a turtlebot heading to the cylinder
• Frontier exploration points
• Keep a path of where we have been? Can keep only last minute of travel

Return to Object (Riley/Jerin)

• Goal: Autonomously return to item(s) requested
• Assumptions: have detected all items requested
• Return to location where object was detected
• Spin if necessary
• Travel to locations in order requested
• Return to home

Spin

• Goal: Function to rotate robot X degrees

Other Considerations

• Reorder supervisor states/logic
• Adjust ASTAR algorithm to eliminate diagonal movements
• Different frontier search methods
• Buffer expand obstacles in occupancy grid (at least for motion planning)