Mobile Robot Control 2021 Group 7: Difference between revisions
TUe\20172067 (talk | contribs) |
TUe\20172067 (talk | contribs) |
||
| Line 16: | Line 16: | ||
=== Evaluation and Reflection === | === Evaluation and Reflection === | ||
During the escape room challenge, the robot behavior was as expected. The feature recognition is designed to localize convex corners and end points of lines, on which the location of the door was based. During the escaperoom challenge the robot first moved to the far end of the room, this was caused by the fact that the wall on the far end of the room was | During the escape room challenge, the robot behavior was as expected. The feature recognition is designed to localize convex corners and end points of lines, on which the location of the door was based. During the escaperoom challenge the robot first moved to the far end of the room, this was caused by the fact that the wall on the far end of the room was too far away to be recognized by the laser range finder. Because of this, two end points where localized by the feature recognition algorithm, making the robot think it has found the exit and drive towards the goal set by the objective calculation. Once the robot arrived at the dead end at the end of the room, it rotated and re-localized a new goal, this time at the correct exit. The robot then managed to drive towards the correct goal and finish the escaperoom challenge well within time. <br> | ||
The robot could have finished sooner with a more robust feature recognition algorithm. For example, a threshold could have been set on the maximum allowed distance between convex corners and line end points, this may have avoided that the robot first drove towards the wall on the far end of the room. The robot would then take more time to rotate and find two points which are within this threshold, improving the localization of the exit. Which ultimately makes the process of exiting the escaperoom more efficient. | The robot could have finished sooner with a more robust feature recognition algorithm. For example, a threshold could have been set on the maximum allowed distance between convex corners and line end points, this may have avoided that the robot first drove towards the wall on the far end of the room. The robot would then take more time to rotate and find two points which are within this threshold, improving the localization of the exit. Which ultimately makes the process of exiting the escaperoom more efficient. | ||
== Hospital Competition == | == Hospital Competition == | ||
Revision as of 15:41, 17 May 2021
Group Members
Tim van Esch - 1235917
Aron Prinsen - 1243943
Thom Samuels - 1267566
Naomi de Vos - 1233610
Joey Wouters - 0813063
Design Document
The Design Document: Media:4SC020_Design_Document_07.pdf
Escape Room Competition
Back-up method
Evaluation and Reflection
During the escape room challenge, the robot behavior was as expected. The feature recognition is designed to localize convex corners and end points of lines, on which the location of the door was based. During the escaperoom challenge the robot first moved to the far end of the room, this was caused by the fact that the wall on the far end of the room was too far away to be recognized by the laser range finder. Because of this, two end points where localized by the feature recognition algorithm, making the robot think it has found the exit and drive towards the goal set by the objective calculation. Once the robot arrived at the dead end at the end of the room, it rotated and re-localized a new goal, this time at the correct exit. The robot then managed to drive towards the correct goal and finish the escaperoom challenge well within time.
The robot could have finished sooner with a more robust feature recognition algorithm. For example, a threshold could have been set on the maximum allowed distance between convex corners and line end points, this may have avoided that the robot first drove towards the wall on the far end of the room. The robot would then take more time to rotate and find two points which are within this threshold, improving the localization of the exit. Which ultimately makes the process of exiting the escaperoom more efficient.