The basic goals the robot was supposed to accomplish were:
Additionally:
Strategy:
Early on, the team decided not to use the A* path planning algorithm to complete the course. Instead a simpler approach was used that involved a lot of wall following. If, at any time, the robot would encounter a wall in the course, it would have to follow that wall until the objective became visible to it again and the path towards the objective is clear. An exception to this is if the robot encountered an unseen weed at any time, even during wall following. In that case the robot would need to perform the weed spraying sequence, then go back on its way to finding the waypoint it was originally headed towards. This was a solid strategy, but generally resulted in slower times than other teams because the robot would need to stay on the walls most of the time and not have a predetermined path for it to go through.
Robot doing the course during the final period:
Trial | Time | Video link |
#1 | -55s | |
#2 | DNF | |
#3 | -105s |
Link to project source code:
https://drive.google.com/file/d/1dcazUtzZdbY8bNc-5P7hoJKvgH7eDOLX/view
Robot performing course weekend before final period:
https://drive.google.com/file/d/1NkqNEtNQHYg7LQtI__PeilVLDkgrZcRe/view
Locating a weed:
Left wall following:
Getting caught in a corner:
Left wall following:
Robot approaches wall Robot aligns itself to follow the wall
Front view of robot after a run:
Final scoreboard:
Full Name: Peter Maneykowski
Major: Systems Engineering and Design
Concentration: Computer Science
Peter particularly enjoyed working in this class because a good amount of C programming was needed to program the robot. He was a big proponent of not using the A* algorithm to complete the course, because he wanted the team to implement their own algorithm to complete the course and requirements.
Peter hopes to find a job in software engineering when he finishes his studies in December.
Full Name: Dante G Nava
Major: Systems Engineering and Design
Concentration: Manufacturing Engineering
Dante enjoyed seeing the robot hit the walls over and over and over again until it didn’t. He thought it fun to see the algorithm evolve as the team integrated functionality and ran into issues.
A graduating senior, Dante intends to start a career in the manufacturing automation industry.
Full Name: Ketaki Tamhankar
Major: Systems Engineering and Design
Concentration: Control Systems
Ketaki particularly enjoyed working with the color vision algorithm, especially developing it into a tool to locate the "weeds."
Ketaki aims to make a difference via bio-inspired engineering, and is spending the summer conducting research on wind turbines.
Name: Samantha Moran
Major: Mechanical Engineering
Samantha most enjoyed puzzling through the state-machine code when new scenarios confused the robot. She is still excited every time it avoids hitting a wall.
Samantha aspires to a career in biomechatronics and is spending the summer working in the medical device industry.
The wall following robot was able to complete the course with a time that rivaled many of the A* robots in the competition. While the robot was not the quickest, this was achieved by ensuring that the algorithm could make its way out of difficult parts of the course without crashing into any obstacles. The team’s weed locating and reporting was also fine tuned to accurately locate all the weeds and report them to Labview. While the wall following algorithm did not always provide the optimal path to the objective, the algorithm allowed it to complete the necessary tasks efficiently. This was only possible after great amounts of testing.
