SAC (2014): RAS Robotics

The images above are the 3 courses each team's robotic car are required to traverse in order to score the maximum amount of points. The trapezoid shown along the bottom of the figure is the side view of the Angle Find course.

Each robotic car is individually timed for each of the 3 courses, and points are awarded based on how fast each robotic car completes the course compared to other robotic cars. There is also a total time awarded to robotic cars that complete all 3 courses. Robotic cars are also awarded points based on the number of courses completed. Prior to the competition, teams are given a chance to score more points by presenting how their car was conceived. Remotely and autonomously controlled robotic cars are both allowed; however, autonomous vehicles are awarded more points.

Testing the Boulder Field at Rowan University

The robotic car needs to meet certain conditions in order to complete each course. All 3 courses are approximately 12 by 4 feet and require the robotic cars to move from the area marked "start" and finish in the area marked "stop". The Line Follow course is a twisty path that makes moving only in a forward direction impossible. The Angle Find course whose main difficultly is its trapezoidal structure. Completing the Angle Find course also requires the robotic vehicle to determine the angle of the trapezoid's incline and then display the angle within 3 degrees of error. The Boulder Field consists of many "boulders", which are square, wooden squares that can have a height up to 1 inch from the ground.

Testing the Line-Following algorithm

And, another important point to note is teams don't compete simultaneously. Some specific rules are also left our of this description for the sake of brevity; a full, detailed explanation of the rules can be found on the competition's website.

Teammate. My teammate with whom I worked was Tal Singer. Please check out his LinkedIn. A bio written by him is included here:

Tal Singer

"My name is Tal Singer, and I am a rising senior majoring in electrical engineering with a focus on computer engineering. Throughout my time at Temple, I have participated in a wide range of leadership, mentorship, and teaching activities. For over two years, I have worked as a teacher’s assistant in the introduction to engineering course and was also a teacher at Temple’s summer robotics camp for high school students. In addition, I have partaken in numerous extracurricular research and leadership projects including the construction of a prosthetic arm that works with the Wii system, robotics competitions, and tutoring in C and C++ languages."

Design. A description on how the robotic car is designed is found on a separate post which can be found here.

Special Thanks. Much thanks goes to Jihun Song, an undergraduate who worked with the team, primarily getting the project's Ledtech seven-segment display to operate over an Arduino UNO (whose source code was easily transferred to a different development board as described under the "Design" page). Much thanks also goes to Rhett Hockenbury, another undergraduate who contributed to the project by cutting out Plexiglass and providing the team hardware for the robotic car's chassis.

Results / Final Thoughts. Overall, our team's robotic car performed as expected. Namely, our robotic car easily surpassed the Line Follow course, doing so only in autonomous mode. The robotic car also managed to get through the Boulder Field course, albeit the robotic car's mode was switched to manual control midway through the course. Unfortunately, due to small size of our robotic car, the robotic car had difficulty traversing the Boulder Field's "boulders" on its own (i.e. autonomously). The issue of the most concern, however, was the Angle Find course.

Top view of the robotic car

A number of our competitions' robotic vehicles were more suited to the Angle Find course than our robotic car. Specifically, the robotic vehicles that were constructed as "miniature tanks" had the surface area and friction needed for climbing the Angle Find course's incline. The "miniature" tank design also appeared to work well with the Boulder Field course; a number of the robotic vehicles that managed to autonomously finish the Angle Find course also managed to autonomously complete the Boulder Field course due to their relatively larger size compared to the "boulders".

Prior to going into the competition, our robotic car was tested to go up roughly 30 degrees. However, the actual degrees of the Angle Find's angle was somewhere around 34 degrees. The reason why we didn't opt to construct our robotic car from a much larger frame was primarily due to limited time in completing the project. Limited resources were also an issue and so nearly all the pieces of the project---including the chassis and most of the electronics---came from either our own supply or from the college.

Jihun song (top-left), Andrew Powell (bottom-left), and Tal Singer (right) working on the project

Considering Tal and I started a little over two weeks before the competition, we are still mostly happy with what we did accomplish. During the presentation before our turn to compete, the judges were pleased about our creative design. In particular, we talked about how we implemented a system in which a button on the Xbox controller would cause the robotic car to "stand up" on its two large wheels. More information on other creative features implemented in the robotic car are found in the Design page, but other features include wireless communication via Xbee transceivers, a GUI constructed in Java, and the capacity to easily switch between manual and autonomous modes with a push on our Xbee controller. If we are to do the same project again, more time allocated to the project and a larger chassis with better friction and lower center of gravity are necessary to improve the project.