Meeting Logs

• Almost completed website by completing meet the team, our approach, and intelligent or not sections. 

• Robot is able to perform a figure-8 on its own and stays within the lines for whole action.

• Completed simulation program of robot solving wumpus world board capable of solving easy and hard boards

• Adapt the physical robot movement code within the algorithm

• Fine-tune some parameters in algorithm 

• Began creating our website. Finished Home page that introduces the rules of the game. 

• Refining simulation program of robot running wumpus world

  • Test algorithm by running generated boards through the program and revising algorithm based on boards that the robot fails to solve

• Robot is able to move in all 4 directions (north, east, south, west) with a signal and adjusts itself to stay in the grid

  • Fixed Bluetooth disconnection issue

  • Further calibrated values for motor power and time on for each action to be more precise

Algorithm

• Fixing some minor logical issues regarding sets operation in Python

• Test extensively on hard puzzles (some puzzles are solved but not updated): https://docs.google.com/document/d/1z9FV1-0fhZOkr7AWpvQTA2q4dT4px8ukQZnOkQp_10Q/edit?usp=sharing

• • Halfway done coding simulating in Python (easier to adapt to physical robot)

    • • Implemented the move back and right logic

      • Can record all the possibilities of items in a list of coordinates

    • TODO:

      • Implement shooting the wumpus (finding what's the logic behind finding the exact position of the wumpus)

      • Testing on more maps

  • Robot can execute functions to move one square forward from a signal

    • Having issues with Bluetooth disconnecting after a few signals

  • Refining our algorithm logic as we test more

Algorithm

• Implemented a way to determine if the robot is in a loop 

• If so, break the robot out of the loop

• • Started coding simulating in Python (easier to adapt to physical robot)

  • Started making python modules for individual movements (move one box forward, turn left, etc)

  • Reconsidering and improving our algorithm logic

  • Perfecting pseudocode for implementation in python

Algorithm

• Implemented finding its way back to the ideal path via minimal distance

• Using moving down and right only for breeze or stench encounter

• Robot stuck in loops sometimes

• • Started simulation in scratch/python (still determining which is a better platform to simulate)

  • The robot is able to travel forward and stop when the sensors detect the white lines of the grid.

  • Reconsidering and improving our algorithm logic

  • Drafting pseudocode for implementation in Python

Algorithm

• The robot will not enter the bottom right quadrant since it will always try to go up, then down, and right if a breeze or stench is felt. 

• We discover that the robot easily gets off the ideal path and gets lost.

Logistics

• Top left is (0,0), bottom right is (3,3)

  • Stored in a 2D list in Python probably

• Orientation: 1, 2, 3, 4=Top, Right, Bottom, Left

• Sending signal: String with a length of 3 of 1 & 0s

  • "111": 1 is True, 0 is False

  • Left: Wumpus

  • Middle: Hole

  • Right: Gold

Algorithm

• Default path of robot: facing up, go straight until corner, then turn right, until (3,1). Since it covers (pass/neighbors) all the possible blocks on board. 

• Retreat strategy (when encounter a stench or breeze): move back 1, move right 1.

Impossible Placements of Holes/Wumpus 

• Either (1,0) or (0,1)

• Both (2,0) and (0,2)

• (1,1) and (Either (2,0) or (0,2))

• TODO

  • Modularize moving forward 1 block, turning right/left/180

  • Coding

  • Testing