For the competition, we created a separate function to complete each individual task and connected each to a button in the GUI. We did this for simplicity and ease of work-flow when programming and testing the tasks. The structure of each function depends on the nature of the task. Tasks one and two utilized a simple state machine that detected the position of a block in the first state, then moved the block to the other side of the board for task one, or onto a stack for task two in the second state. If the the first state determined that all blocks were in the correct position, it would instead transition to a third and final end state.
Tasks three and four were somewhat more complicated, as they must be prepared for cases where the next block that should be placed in the line or stack were not visible. Here we define "caching" a block as placing it in a position that it not its ultimate goal, in order to uncover other blocks.…show more content… 21 blocks were placed 16cm away from the arm with the flat side of the block facing the center of the board, 21 more were placed at 16cm with the corner of the block facing the center. The next 21 were placed at 28.5cm with the flat side facing the arm, and a final 21 were placed at 28.5cm with the corner facing the arm. Out of these trials, all 42 blocks were successfully picked up at 16cm. Out of the 28.5cm blocks, the arm failed to pick up 3 of them when the flat side faced the robot, and failed to pick up 9 of them when the corner faced the robot. These cases in which the gripper performed relatively poorly were at the edge of the range of the arm, and we believe they may be the worst case position for a cube while still being reachable by the arm. In the competition, we never had to pick up a block in these conditions, and we believed the performance to be acceptable given the likelihood of
